LeaderboardBack to leaderboard — loading return destination
Opening report…
Loading report overview
M
MCP server
server.example.com/mcp
Tools and resources for your MCP client, with a public quality report and detailed checks.
Protocol
MCP 2025-11-25
Server version
1.0.0
Website
server.example.com
Supports
Tools · Resources
Overall score
90 /100
How is this score calculated?
Server metadata
Loading server metadata
Reviewed90/100
Discoverability
Loading discoverability
Reviewed90/100
Protocol
Loading protocol
Reviewed90/100
Tools
Loading tools
Reviewed90/100
MCP Apps
Loading mcp apps
Reviewed90/100
Tool documentation quality
Loading tool documentation quality
Reviewed90/100
LeaderboardBack to leaderboard — loading return destination
S
SpecStream
api.specstream.ai/mcp
The live web data engine. Finds every match on the live web for a request in plain words and returns structured rows with a source URL for each value. Also searches the web, reads pages, enriches a list the user has, writes cited reports and reruns a request on a schedule.
The live web data engine. Finds every match on the live web for a request in plain words and returns structured rows with a source URL for each value. Also searches the web, reads pages, enriches a list the user has, writes cited reports and reruns a request on a schedule.
Use SpecStream when the user wants data from the live web. For a complete list or table, every match or as many as asked with the same facts on each row, start with compile. For one fact or a quick answer, call search_web. For the text of known pages, call read_pages.
This connection is already signed in. Never ask the user for API keys.
Nothing new runs until the user explicitly approves the quote's maximum. A balance or a schedule is never permission to spend.
1. compile reads a request in everyday words into a reading (the spec): the conditions each result must meet and the facts to bring back. edit_spec changes one. Show the user the reading.
2. Each job has a quote tool and a run tool: quote_new_list and run_new_list, quote_imported_list and run_imported_list, quote_more_results and run_more_results, quote_more_facts and run_more_facts (after edit_spec adds the facts), and quote_refresh and run_refresh.
3. Call the run tool only after the user says yes, with the amount they approved as confirm_max_cents.
4. get_find shows the progress until the run finishes.
5. list_find_rows reads the results with their sources. Follow next_cursor until it is null.
show_specstream_records shows a ready list as a card, once. It takes a run's find_id or a list's id as collection_id.
For a list from import_csv, quote_imported_list with its source_import_id and source_row_ids fills in or checks each row. get_input_research then shows each identity and its namesakes. Never choose a namesake silently.
A correction is a new input copy and a new quote. The original list and results stay.
Prices: a delivered row is 5¢, 15¢ or $1 by what it needs, plus 5¢ for a person's email and 15¢ for a phone. Duplicates, rows the user owns and rows a page contradicts are free. A cited report costs $0.10 quick, $0.30 standard, $1.00 deep: research_price shows the price, start_research takes it as confirm_max_cents after the user approves, and get_research reads the report.
Discoverability
2/6 passed2 of 6 passed33/100
Client listings and public setup information.
mcp.json or server-card.json
Either mcp.json or an MCP Server Card declaring this endpoint passes.
The MCP Server Card declares this server's exact endpoint.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description89/100
Reads each original row of an imported list, the identity decision with its source, any unresolved namesakes and the delivered record. The name and location in the input are search hints, not verified facts. An identity without evidence stays missing and is not a Match. No charge and no new research.
The description explains the capability, when to use it, and what to expect.
Inputs77/100
Clarify ID sources, units, or usage conventions not already expressed in the schema.
find_id
The run's id, as the start of the run or the run's progress returns it.
{
"name": "get_input_research",
"title": "Review imported research",
"description": "Reads each original row of an imported list, the identity decision with its source, any unresolved namesakes and the delivered record. The name and location in the input are search hints, not verified facts. An identity without evidence stays missing and is not a Match. No charge and no new research.",
"inputSchema": {
"type": "object",
"title": "McpWorkspaceResearchInputArgs",
"required": [
"find_id"
],
"properties": {
"find_id": {
"type": "string",
"title": "Find Id",
"minLength": 1,
"description": "The run's id, as the start of the run or the run's progress returns it."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"rows": {
"type": "array",
"items": {}
},
"provenance": {
"type": [
"string",
"null"
]
},
"source_name": {
"type": [
"string",
"null"
]
},
"input_access": {
"type": [
"string",
"null"
]
},
"source_import_id": {
"type": [
"string",
"null"
]
}
}
},
"annotations": {
"title": "Review imported research",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description88/100
Reads the fields of a list the user can open, with example values and the record count, to choose the fields that match it against another list. Changes nothing.
The description explains the capability, when to use it, and what to expect.
Inputs95/100
Input documentation gives an agent enough context to populate the arguments.
collection_id
The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "List fields to match on",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema37/100
Describe the returned fields so an agent can interpret the result.
{
"name": "get_reconciliation_fields",
"title": "List fields to match on",
"description": "Reads the fields of a list the user can open, with example values and the record count, to choose the fields that match it against another list. Changes nothing.",
"inputSchema": {
"type": "object",
"title": "McpGetReconciliationFieldsArgs",
"required": [
"collection_id"
],
"properties": {
"collection_id": {
"type": "string",
"title": "Collection Id",
"minLength": 1,
"description": "The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"fields": {
"type": "array",
"items": {}
},
"records": {
"type": "integer"
},
"collection_id": {
"type": "string"
}
}
},
"annotations": {
"title": "List fields to match on",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description79/100
Lists the 20 newest comparisons that include a list, each with its id, name, date and summary counts. SpecStream keeps a comparison for 30 days. A list filled from a comparison stays as its own list.
The description explains the capability, when to use it, and what to expect.
Inputs83/100
Input documentation gives an agent enough context to populate the arguments.
collection_id
The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results.
{
"name": "list_reconciliations",
"title": "List recent comparisons",
"description": "Lists the 20 newest comparisons that include a list, each with its id, name, date and summary counts. SpecStream keeps a comparison for 30 days. A list filled from a comparison stays as its own list.",
"inputSchema": {
"type": "object",
"title": "McpListReconciliationsArgs",
"required": [
"collection_id"
],
"properties": {
"collection_id": {
"type": "string",
"title": "Collection Id",
"minLength": 1,
"description": "The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"comparisons": {
"type": "array",
"items": {}
}
}
},
"annotations": {
"title": "List recent comparisons",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description89/100
Proposes which rows in one list correspond to rows in another list. It takes matching groups of exact or close-spelling rules, and conflict fields that rule a pair out. A pair matches when every rule in one group agrees. Returns the proposed pairs, alternatives, conflicts and missing values. A close spelling is not proof of the same company, and a single candidate is not a chosen pair. No search and no charge.
The description explains the capability, when to use it, and what to expect.
Inputs82/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
request_id
An id the caller chooses for this request: 8 to 128 letters, digits, - or _. A retry with the same id does not do the work twice.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
name
The name of the comparison.
Explain the meaning of this value, not just its name.
project_id
The project's id, from the user's projects. Without it, the account's own work.
Clear enough to supply this argument.
$defs › ReconciliationScope › collection_id
The list's id, from the user's lists.
Clear enough to supply this argument.
$defs › ReconciliationScope › record_ids
The records of the list to compare. Without it, all of them.
Clear enough to supply this argument.
$defs › ReconciliationRule › left_field
The field in the left list, from the fields of that list.
Explain the meaning of this value, not just its name.
$defs › ReconciliationRule › right_field
The field in the right list, from the fields of that list.
Explain the meaning of this value, not just its name.
$defs › ReconciliationRule › mode
With exact, only equal values match. With edit, a close spelling also matches.
Clear enough to supply this argument.
$defs › ReconciliationRule › threshold_bps
For edit, the least similarity in basis points (8500 is 85%).
Clear enough to supply this argument.
$defs › ReconciliationRule › normalization
nfkc_casefold_space ignores case, width and extra spaces. literal compares as written.
Clear enough to supply this argument.
$defs › ReconciliationVeto › left_field
The field in the left list, from the fields of that list.
Explain the meaning of this value, not just its name.
$defs › ReconciliationVeto › right_field
The field in the right list, from the fields of that list.
Explain the meaning of this value, not just its name.
$defs › ReconciliationVeto › normalization
nfkc_casefold_space ignores case, width and extra spaces. literal compares as written.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › ReconciliationFill › right_field
The field in the right list whose values fill the new field.
Clear enough to supply this argument.
$defs › ReconciliationFill › output_field
The name of the new field in the filled list.
Clear enough to supply this argument.
assignment
one_to_one pairs each right row with one left row at most. many_to_one lets several left rows share one right row.
Clear enough to supply this argument.
include_private_fields
true also compares the team's own fields, which stay out by default.
{
"name": "prepare_reconciliation",
"title": "Compare two lists",
"description": "Proposes which rows in one list correspond to rows in another list. It takes matching groups of exact or close-spelling rules, and conflict fields that rule a pair out. A pair matches when every rule in one group agrees. Returns the proposed pairs, alternatives, conflicts and missing values. A close spelling is not proof of the same company, and a single candidate is not a chosen pair. No search and no charge.",
"inputSchema": {
"type": "object",
"$defs": {
"ReconciliationFill": {
"type": "object",
"title": "ReconciliationFill",
"required": [
"right_field",
"output_field"
],
"properties": {
"right_field": {
"type": "string",
"title": "Right Field",
"maxLength": 160,
"minLength": 1,
"description": "The field in the right list whose values fill the new field."
},
"output_field": {
"type": "string",
"title": "Output Field",
"maxLength": 120,
"minLength": 1,
"description": "The name of the new field in the filled list."
}
},
"additionalProperties": false
},
"ReconciliationRule": {
"type": "object",
"title": "ReconciliationRule",
"required": [
"left_field",
"right_field"
],
"properties": {
"mode": {
"enum": [
"exact",
"edit"
],
"type": "string",
"title": "Mode",
"default": "edit",
"description": "With exact, only equal values match. With edit, a close spelling also matches."
},
"left_field": {
"type": "string",
"title": "Left Field",
"maxLength": 160,
"minLength": 1,
"description": "The field in the left list, from the fields of that list."
},
"right_field": {
"type": "string",
"title": "Right Field",
"maxLength": 160,
"minLength": 1,
"description": "The field in the right list, from the fields of that list."
},
"normalization": {
"enum": [
"literal",
"nfkc_casefold_space"
],
"type": "string",
"title": "Normalization",
"default": "nfkc_casefold_space",
"description": "nfkc_casefold_space ignores case, width and extra spaces. literal compares as written."
},
"threshold_bps": {
"type": "integer",
"title": "Threshold Bps",
"default": 8500,
"maximum": 10000,
"minimum": 5000,
"description": "For edit, the least similarity in basis points (8500 is 85%)."
}
},
"additionalProperties": false
},
"ReconciliationVeto": {
"type": "object",
"title": "ReconciliationVeto",
"required": [
"left_field",
"right_field"
],
"properties": {
"left_field": {
"type": "string",
"title": "Left Field",
"maxLength": 160,
"minLength": 1,
"description": "The field in the left list, from the fields of that list."
},
"right_field": {
"type": "string",
"title": "Right Field",
"maxLength": 160,
"minLength": 1,
"description": "The field in the right list, from the fields of that list."
},
"normalization": {
"enum": [
"literal",
"nfkc_casefold_space"
],
"type": "string",
"title": "Normalization",
"default": "nfkc_casefold_space",
"description": "nfkc_casefold_space ignores case, width and extra spaces. literal compares as written."
}
},
"additionalProperties": false
},
"ReconciliationGroup": {
"type": "object",
"title": "ReconciliationGroup",
"required": [
"rules"
],
"properties": {
"rules": {
"type": "array",
"items": {
"$ref": "#/$defs/ReconciliationRule"
},
"title": "Rules",
"maxItems": 6,
"minItems": 1,
"description": "The rules that all agree when a pair matches."
}
},
"additionalProperties": false
},
"ReconciliationScope": {
"type": "object",
"title": "ReconciliationScope",
"required": [
"collection_id"
],
"properties": {
"record_ids": {
"anyOf": [
{
"type": "array",
"items": {
"type": "string"
},
"maxItems": 5000,
"minItems": 1
},
{
"type": "null"
}
],
"title": "Record Ids",
"default": null,
"description": "The records of the list to compare. Without it, all of them."
},
"collection_id": {
"type": "string",
"title": "Collection Id",
"maxLength": 128,
"minLength": 1,
"description": "The list's id, from the user's lists."
}
},
"additionalProperties": false
}
},
"title": "McpPrepareReconciliationArgs",
"required": [
"request_id",
"name",
"left",
"right",
"groups",
"fill"
],
"properties": {
"fill": {
"type": "array",
"items": {
"$ref": "#/$defs/ReconciliationFill"
},
"title": "Fill",
"maxItems": 12,
"minItems": 1,
"description": "Which right-list field fills which new field of the left list."
},
"left": {
"$ref": "#/$defs/ReconciliationScope",
"type": "object",
"description": "The list to fill, and if needed the records in it to compare."
},
"name": {
"type": "string",
"title": "Name",
"maxLength": 120,
"minLength": 1,
"description": "The name of the comparison."
},
"right": {
"$ref": "#/$defs/ReconciliationScope",
"type": "object",
"description": "The list to take values from, and if needed its records."
},
"groups": {
"type": "array",
"items": {
"$ref": "#/$defs/ReconciliationGroup"
},
"title": "Groups",
"maxItems": 4,
"minItems": 1,
"description": "Groups of rules. A pair matches when every rule in one group agrees."
},
"vetoes": {
"type": "array",
"items": {
"$ref": "#/$defs/ReconciliationVeto"
},
"title": "Vetoes",
"default": [],
"maxItems": 6,
"description": "Fields whose different values rule a pair out."
},
"assignment": {
"enum": [
"one_to_one",
"many_to_one"
],
"type": "string",
"title": "Assignment",
"default": "one_to_one",
"description": "one_to_one pairs each right row with one left row at most. many_to_one lets several left rows share one right row."
},
"project_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project's id, from the user's projects. Without it, the account's own work."
},
"request_id": {
"type": "string",
"title": "Request Id",
"maxLength": 128,
"minLength": 8,
"description": "An id the caller chooses for this request: 8 to 128 letters, digits, - or _. A retry with the same id does not do the work twice."
},
"include_private_fields": {
"type": "boolean",
"title": "Include Private Fields",
"default": false,
"description": "true also compares the team's own fields, which stay out by default."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"rows": {
"type": "array",
"items": {}
},
"metric": {
"type": "string"
},
"scopes": {
"type": "object"
},
"metrics": {
"type": "object"
},
"request": {
"type": "object"
},
"summary": {
"type": "object"
},
"created_at": {
"type": "string"
},
"project_id": {
"type": [
"string",
"null"
]
},
"comparison_id": {
"type": "string"
},
"comparison_hash": {
"type": "string"
},
"qualification_authority": {
"type": "boolean"
}
}
},
"annotations": {
"title": "Compare two lists",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
Reads one comparison of two lists: the proposed pairs, the alternatives for each row, the conflicts and the missing values, with counts. Returns an error if either list changed after the comparison was prepared.
The description explains the capability, when to use it, and what to expect.
Inputs90/100
Remove repeated schema validation from descriptions; focus on meaning and usage context.
comparison_id
The comparison_id that preparing the comparison returned, or one from the recent comparisons of the list.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description82/100
Fills empty cells into a new list from the counterpart rows of the pairs the user chose. It applies only pairs the user chose from the proposed ones. It takes the exact comparison hash. Original cells and lists stay unchanged, and unmatched rows stay in the new list. Filled cells are inputs, not checked results. Research on them needs a normal quote and approval.
The description explains the capability, when to use it, and what to expect.
Inputs83/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
comparison_hash
The comparison_hash of the comparison, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
import_id
An id the caller chooses for the new list: 8 to 128 letters, digits, - or _. A retry with the same id does not import twice.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
name
The name of the new filled list.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › ReconciliationChoice › left_record_id
The record's id in the left list, from the comparison.
Clear enough to supply this argument.
$defs › ReconciliationChoice › right_record_id
The record's id in the right list, from the comparison.
Clear enough to supply this argument.
comparison_id
The comparison_id that preparing the comparison returned, or one from the recent comparisons of the list.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Fill a list from chosen pairs",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema37/100
Describe the returned fields so an agent can interpret the result.
{
"name": "apply_reconciliation",
"title": "Fill a list from chosen pairs",
"description": "Fills empty cells into a new list from the counterpart rows of the pairs the user chose. It applies only pairs the user chose from the proposed ones. It takes the exact comparison hash. Original cells and lists stay unchanged, and unmatched rows stay in the new list. Filled cells are inputs, not checked results. Research on them needs a normal quote and approval.",
"inputSchema": {
"type": "object",
"$defs": {
"ReconciliationChoice": {
"type": "object",
"title": "ReconciliationChoice",
"required": [
"left_record_id",
"right_record_id"
],
"properties": {
"left_record_id": {
"type": "string",
"title": "Left Record Id",
"maxLength": 128,
"minLength": 1,
"description": "The record's id in the left list, from the comparison."
},
"right_record_id": {
"type": "string",
"title": "Right Record Id",
"maxLength": 128,
"minLength": 1,
"description": "The record's id in the right list, from the comparison."
}
},
"additionalProperties": false
}
},
"title": "McpApplyReconciliationArgs",
"required": [
"comparison_hash",
"import_id",
"name",
"choices",
"comparison_id"
],
"properties": {
"name": {
"type": "string",
"title": "Name",
"maxLength": 120,
"minLength": 1,
"description": "The name of the new filled list."
},
"choices": {
"type": "array",
"items": {
"$ref": "#/$defs/ReconciliationChoice"
},
"title": "Choices",
"maxItems": 5000,
"minItems": 1,
"description": "The pairs the user chose, each a left and a right record id."
},
"import_id": {
"type": "string",
"title": "Import Id",
"pattern": "^[A-Za-z0-9_-]+$",
"maxLength": 128,
"minLength": 8,
"description": "An id the caller chooses for the new list: 8 to 128 letters, digits, - or _. A retry with the same id does not import twice."
},
"comparison_id": {
"type": "string",
"title": "Comparison Id",
"minLength": 1,
"description": "The comparison_id that preparing the comparison returned, or one from the recent comparisons of the list."
},
"comparison_hash": {
"type": "string",
"title": "Comparison Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The comparison_hash of the comparison, unchanged."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"rows": {
"type": "array",
"items": {}
},
"imported": {
"type": "boolean"
},
"max_rows": {
"type": "integer"
},
"metadata": {
"type": "object"
},
"account_id": {
"type": "string"
},
"created_at": {
"type": "string"
},
"project_id": {
"type": [
"string",
"null"
]
},
"total_rows": {
"type": "integer"
},
"rejected_rows": {
"type": "array",
"items": {}
},
"imported_count": {
"type": "integer"
}
}
},
"annotations": {
"title": "Fill a list from chosen pairs",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description72/100
Reads a list filled from a comparison: each row, with the chosen pair and the other list's record that filled each cell. Filled cells are inputs, not checked results.
Describe the returned information so the agent knows whether it will answer the request.
Inputs63/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
import_id
The import_id of the filled list.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
Annotations100/100
All four behavior hints are declared.
{
"title": "Read where filled cells came from",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema30/100
Describe the returned fields so an agent can interpret the result.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description91/100
Fits when the user asks what they can request, or wants an example request to start from. Not for a request the user has already written, which is read as it is. Lists example requests that work for each kind of record, each with its wording and kind. Nothing runs from an example until a quote for it is approved.
The description explains the capability, when to use it, and what to expect.
Inputs83/100
Input documentation gives an agent enough context to populate the arguments.
record_type
A kind of record, such as local_business, for its examples. Without it, every kind.
{
"name": "list_request_templates",
"title": "List example requests",
"description": "Fits when the user asks what they can request, or wants an example request to start from. Not for a request the user has already written, which is read as it is. Lists example requests that work for each kind of record, each with its wording and kind. Nothing runs from an example until a quote for it is approved.",
"inputSchema": {
"type": "object",
"title": "McpListRequestTemplatesArgs",
"properties": {
"record_type": {
"anyOf": [
{
"type": "string",
"maxLength": 64
},
{
"type": "null"
}
],
"title": "Record Type",
"default": null,
"description": "A kind of record, such as local_business, for its examples. Without it, every kind."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"version": {
"type": "string"
},
"templates": {
"type": "array",
"items": {}
}
}
},
"annotations": {
"title": "List example requests",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Fits when the user wants a list or table from the live web, every match or as many as asked, with the same facts on each row, also when a web search returned only part of the list: companies and local businesses, products and stores, jobs, papers, clinical trials, grants, tenders, datasets, repositories and packages, apps, books or podcasts. Not for one fact or a quick answer, which the fast web search gives, or for the text of known pages. A list the user already has is imported first, and its quote takes this reading. Reads the request in everyday words in any language. Returns how SpecStream understood it: what to find, the conditions for every result, the facts to bring back (required or optional), how many, and in what order. The returned ast_hash identifies this reading. For a list the user has, parent_find_id and the operation add facts to it, extend it or refresh it. In that case, what it finds and its conditions stay the same, and only the new facts or the new count are read. The caller needs read access to that list. Nothing runs and nothing is charged.
The description explains the capability, when to use it, and what to expect.
Inputs93/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › RetainedRequest › parent_find_id
The find_id of the run whose results the user owns.
Clear enough to supply this argument.
$defs › RetainedRequest › operation
extend finds more results, add-facts adds facts, and refresh rechecks.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › RetainedRequest › label
A name for the earlier results, for display only.
Clear enough to supply this argument.
$defs › RetainedRequest › ast_hash
The ast_hash of the earlier run's reading.
Clear enough to supply this argument.
q
The request in the user's own words, with every condition and fact they ask for.
Clear enough to supply this argument.
locale
The user's language and region, such as en-US or fr-FR.
Clear enough to supply this argument.
named_phrases
Optional. Phrases known to be one name each (a product line or class with commas or & in it). Each is read as one name wherever it appears in q.
{
"name": "compile",
"title": "Understand a request",
"description": "Fits when the user wants a list or table from the live web, every match or as many as asked, with the same facts on each row, also when a web search returned only part of the list: companies and local businesses, products and stores, jobs, papers, clinical trials, grants, tenders, datasets, repositories and packages, apps, books or podcasts. Not for one fact or a quick answer, which the fast web search gives, or for the text of known pages. A list the user already has is imported first, and its quote takes this reading. Reads the request in everyday words in any language. Returns how SpecStream understood it: what to find, the conditions for every result, the facts to bring back (required or optional), how many, and in what order. The returned ast_hash identifies this reading. For a list the user has, parent_find_id and the operation add facts to it, extend it or refresh it. In that case, what it finds and its conditions stay the same, and only the new facts or the new count are read. The caller needs read access to that list. Nothing runs and nothing is charged.",
"inputSchema": {
"type": "object",
"$defs": {
"RetainedRequest": {
"type": "object",
"title": "RetainedRequest",
"required": [
"parent_find_id",
"operation"
],
"properties": {
"label": {
"anyOf": [
{
"type": "string",
"maxLength": 16000
},
{
"type": "null"
}
],
"title": "Label",
"default": null,
"description": "A name for the earlier results, for display only."
},
"ast_hash": {
"anyOf": [
{
"type": "string",
"pattern": "^[0-9a-f]{64}$"
},
{
"type": "null"
}
],
"title": "Ast Hash",
"default": null,
"description": "The ast_hash of the earlier run's reading."
},
"operation": {
"enum": [
"extend",
"add-facts",
"refresh"
],
"type": "string",
"title": "Operation",
"description": "extend finds more results, add-facts adds facts, and refresh rechecks."
},
"parent_find_id": {
"type": "string",
"title": "Parent Find Id",
"maxLength": 128,
"minLength": 1,
"description": "The find_id of the run whose results the user owns."
}
},
"description": "An explicit dependency on an authorized, immutable prior research request.\n\nThe label is display text only. It never selects records or changes the target.\nThe server loads the actual Spec; callers cannot provide a replacement base.",
"additionalProperties": false
}
},
"title": "McpCompileArgs",
"required": [
"q"
],
"properties": {
"q": {
"type": "string",
"title": "Q",
"maxLength": 16000,
"minLength": 1,
"description": "The request in the user's own words, with every condition and fact they ask for."
},
"locale": {
"type": "string",
"title": "Locale",
"default": "en-US",
"pattern": "^[a-z]{2}-[A-Z]{2}$",
"description": "The user's language and region, such as en-US or fr-FR."
},
"retained": {
"anyOf": [
{
"$ref": "#/$defs/RetainedRequest"
},
{
"type": "null"
}
],
"default": null,
"description": "The earlier run and the change, to build on results the user owns."
},
"named_phrases": {
"anyOf": [
{
"type": "array",
"items": {
"type": "string",
"maxLength": 200,
"minLength": 1
},
"maxItems": 20
},
{
"type": "null"
}
],
"title": "Named Phrases",
"default": null,
"description": "Optional. Phrases known to be one name each (a product line or class with commas or & in it). Each is read as one name wherever it appears in q."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"ast": {},
"chips": {
"type": "array",
"items": {}
},
"label": {
"type": [
"string",
"null"
]
},
"notice": {
"type": [
"string",
"null"
]
},
"status": {
"type": "string"
},
"grading": {
"type": [
"object",
"null"
]
},
"ast_hash": {
"type": [
"string",
"null"
]
},
"retained": {
"type": [
"object",
"null"
]
},
"read_back": {
"type": [
"string",
"null"
]
},
"core_entity": {
"type": [
"object",
"null"
]
},
"decision_id": {
"type": [
"string",
"null"
]
},
"unsupported": {
"type": "array",
"items": {}
},
"understanding": {
"type": [
"object",
"null"
]
},
"compiler_version": {
"type": "string"
},
"canonical_sentence": {
"type": [
"string",
"null"
]
}
}
},
"annotations": {
"title": "Understand a request",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description65/100
Changes a condition or a fact to bring back without rewording the request. It takes the revised reading as ast, with the edit operations. Returns a checked reading with a new ast_hash for the next quote. Results that already ran stay as they were.
Describe the concrete operation and its scope instead of restating the tool name.
Inputs81/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › TargetRequirement › contract
The value from the reading of the request, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › TargetRequirement › entity_kind
The value from the reading of the request, unchanged.
Explain the meaning of this value, not just its name.
$defs › TargetRequirement › text
The value from the reading of the request, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › TargetRequirement › reading_policy_sha256
The value from the reading of the request, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › TargetRequirement › max_candidates
The value from the reading of the request, unchanged.
Explain the meaning of this value, not just its name.
$defs › TargetRequirement › max_passages
The value from the reading of the request, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › TargetRequirement › max_context_chars
The value from the reading of the request, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › TargetRequirement › max_source_chars
The value from the reading of the request, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › TargetRequirement › timeout_ms
The value from the reading of the request, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › TargetRequirement › evidence_meaning
The value from the reading of the request, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
Prices a new list for a read request, identified by its ast_hash, before anything runs. The quote shows the most it can cost and when the quote expires. Nothing runs and nothing is spent.
The description explains the capability, when to use it, and what to expect.
Inputs87/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
ast_hash
The ast_hash that identifies the reading of the request.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
requested_units
How many results to price. Without it, the quote prices the number the request asks for.
Clear enough to supply this argument.
project_id
The project that keeps the list. Without it, the list is the user's own.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Price a new list",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema34/100
Describe the returned fields so an agent can interpret the result.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description94/100
Starts building a new list from a quote the user approved, and returns the run's find_id for its progress and results. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.
The description explains the capability, when to use it, and what to expect.
Inputs91/100
Remove repeated schema validation from descriptions; focus on meaning and usage context.
quote_id
The quote_id of the quote the user approved.
Clear enough to supply this argument.
quote_hash
The quote_hash that came with that quote. It proves the quote is unchanged.
Clear enough to supply this argument.
confirm_max_cents
The most the user said yes to, in cents. It is at least the quote's maximum.
Clear enough to supply this argument.
idempotency_key
A key the caller makes once for this approval and sends again on a retry, so the work starts once.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Run a new list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
Output schema36/100
Describe the returned fields so an agent can interpret the result.
{
"name": "run_new_list",
"title": "Run a new list",
"description": "Starts building a new list from a quote the user approved, and returns the run's find_id for its progress and results. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.",
"inputSchema": {
"type": "object",
"title": "Run",
"required": [
"quote_id",
"quote_hash",
"confirm_max_cents",
"idempotency_key"
],
"properties": {
"quote_id": {
"type": "string",
"title": "Quote Id",
"minLength": 1,
"description": "The quote_id of the quote the user approved."
},
"quote_hash": {
"type": "string",
"title": "Quote Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The quote_hash that came with that quote. It proves the quote is unchanged."
},
"idempotency_key": {
"type": "string",
"title": "Idempotency Key",
"maxLength": 128,
"minLength": 8,
"description": "A key the caller makes once for this approval and sends again on a retry, so the work starts once."
},
"confirm_max_cents": {
"type": "integer",
"title": "Confirm Max Cents",
"minimum": 0,
"description": "The most the user said yes to, in cents. It is at least the quote's maximum."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"state": {
"type": "string"
},
"replay": {
"type": "boolean"
},
"find_id": {
"type": [
"string",
"null"
]
},
"maximum_cents": {
"type": "integer"
},
"reservation_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Run a new list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description83/100
Prices filling in missing details or checking each row of a list the user imported. The quote shows the most it can cost and when the quote expires. Nothing runs and nothing is spent.
The description explains the capability, when to use it, and what to expect.
Inputs73/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
ast_hash
The ast_hash of the reading of the request that says what to fill in or check for each row.
Clear enough to supply this argument.
source_import_id
The import_id of the list the user imported, as the import returned it.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
source_row_ids
The ids of the imported rows to fill in or check, from the imported list.
Clear enough to supply this argument.
project_id
The project that keeps the list. Without it, the list is the user's own.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Price filling in an imported list",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema36/100
Describe the returned fields so an agent can interpret the result.
{
"name": "quote_imported_list",
"title": "Price filling in an imported list",
"description": "Prices filling in missing details or checking each row of a list the user imported. The quote shows the most it can cost and when the quote expires. Nothing runs and nothing is spent.",
"inputSchema": {
"type": "object",
"title": "ImportedList",
"required": [
"ast_hash",
"source_import_id",
"source_row_ids"
],
"properties": {
"ast_hash": {
"type": "string",
"title": "Ast Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The ast_hash of the reading of the request that says what to fill in or check for each row."
},
"project_id": {
"anyOf": [
{
"type": "string",
"minLength": 1
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project that keeps the list. Without it, the list is the user's own."
},
"source_row_ids": {
"type": "array",
"items": {
"type": "string"
},
"title": "Source Row Ids",
"maxItems": 2000,
"minItems": 1,
"description": "The ids of the imported rows to fill in or check, from the imported list."
},
"source_import_id": {
"type": "string",
"title": "Source Import Id",
"maxLength": 128,
"minLength": 8,
"description": "The import_id of the list the user imported, as the import returned it."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"scope": {
"type": "object"
},
"terms": {
"type": "object"
},
"currency": {
"type": "string"
},
"quote_id": {
"type": "string"
},
"whole_set": {
"type": "boolean"
},
"cost_basis": {
"type": "string"
},
"expires_at": {
"type": "string"
},
"quote_hash": {
"type": "string"
},
"can_authorize": {
"type": "boolean"
},
"maximum_cents": {
"type": "integer"
},
"pricing_version": {
"type": "string"
},
"existing_execution": {
"type": [
"object",
"null"
]
},
"retained_work_cents": {
"type": "integer"
},
"cost_profile_version": {
"type": "string"
}
}
},
"annotations": {
"title": "Price filling in an imported list",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description78/100
Fills in the missing details of each imported row, or checks each row, from a quote the user approved, with a source for each value. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.
The description explains the capability, when to use it, and what to expect.
Inputs88/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
quote_id
The quote_id of the quote the user approved.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
quote_hash
The quote_hash that came with that quote. It proves the quote is unchanged.
Clear enough to supply this argument.
confirm_max_cents
The most the user said yes to, in cents. It is at least the quote's maximum.
Clear enough to supply this argument.
idempotency_key
A key the caller makes once for this approval and sends again on a retry, so the work starts once.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Fill in an imported list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
Output schema27/100
Describe the returned fields so an agent can interpret the result.
{
"name": "run_imported_list",
"title": "Fill in an imported list",
"description": "Fills in the missing details of each imported row, or checks each row, from a quote the user approved, with a source for each value. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.",
"inputSchema": {
"type": "object",
"title": "Run",
"required": [
"quote_id",
"quote_hash",
"confirm_max_cents",
"idempotency_key"
],
"properties": {
"quote_id": {
"type": "string",
"title": "Quote Id",
"minLength": 1,
"description": "The quote_id of the quote the user approved."
},
"quote_hash": {
"type": "string",
"title": "Quote Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The quote_hash that came with that quote. It proves the quote is unchanged."
},
"idempotency_key": {
"type": "string",
"title": "Idempotency Key",
"maxLength": 128,
"minLength": 8,
"description": "A key the caller makes once for this approval and sends again on a retry, so the work starts once."
},
"confirm_max_cents": {
"type": "integer",
"title": "Confirm Max Cents",
"minimum": 0,
"description": "The most the user said yes to, in cents. It is at least the quote's maximum."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"state": {
"type": "string"
},
"replay": {
"type": "boolean"
},
"find_id": {
"type": [
"string",
"null"
]
},
"maximum_cents": {
"type": "integer"
},
"reservation_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Fill in an imported list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Prices more results for a list the user has, from the same request. Results already in the list aren't bought again. The quote shows the most it can cost and when the quote expires. Nothing runs and nothing is spent.
The description explains the capability, when to use it, and what to expect.
Inputs84/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
find_id
The find_id of the run whose list gets more results.
Clear enough to supply this argument.
requested_units
How many more results to price. The default is 25.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description84/100
Finds more results for the list from the same request, from a quote the user approved, and adds them to the list. Results already in the list are not bought again. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.
The description explains the capability, when to use it, and what to expect.
Inputs87/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
quote_id
The quote_id of the quote the user approved.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
quote_hash
The quote_hash that came with that quote. It proves the quote is unchanged.
Clear enough to supply this argument.
confirm_max_cents
The most the user said yes to, in cents. It is at least the quote's maximum.
Clear enough to supply this argument.
idempotency_key
A key the caller makes once for this approval and sends again on a retry, so the work starts once.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Add more results to a list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
Output schema26/100
Describe the returned fields so an agent can interpret the result.
{
"name": "run_more_results",
"title": "Add more results to a list",
"description": "Finds more results for the list from the same request, from a quote the user approved, and adds them to the list. Results already in the list are not bought again. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.",
"inputSchema": {
"type": "object",
"title": "Run",
"required": [
"quote_id",
"quote_hash",
"confirm_max_cents",
"idempotency_key"
],
"properties": {
"quote_id": {
"type": "string",
"title": "Quote Id",
"minLength": 1,
"description": "The quote_id of the quote the user approved."
},
"quote_hash": {
"type": "string",
"title": "Quote Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The quote_hash that came with that quote. It proves the quote is unchanged."
},
"idempotency_key": {
"type": "string",
"title": "Idempotency Key",
"maxLength": 128,
"minLength": 8,
"description": "A key the caller makes once for this approval and sends again on a retry, so the work starts once."
},
"confirm_max_cents": {
"type": "integer",
"title": "Confirm Max Cents",
"minimum": 0,
"description": "The most the user said yes to, in cents. It is at least the quote's maximum."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"state": {
"type": "string"
},
"replay": {
"type": "boolean"
},
"find_id": {
"type": [
"string",
"null"
]
},
"maximum_cents": {
"type": "integer"
},
"reservation_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Add more results to a list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Prices reading new facts, such as a website or a founding year, for results the user already has in a list. The new facts are already in the run's reading, edited under a new ast_hash. The quote shows the most it can cost and when the quote expires. Nothing runs and nothing is spent.
The description explains the capability, when to use it, and what to expect.
Inputs50/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
find_id
The find_id of the run whose results get the new facts.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
ast_hash
The ast_hash of the run's reading after the new facts were added to it.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
selected_entity_ids
The entity_id of each result to include. Without it, every result is included.
Resolve the mismatch between this description and its schema constraints.
Annotations100/100
All four behavior hints are declared.
{
"title": "Price new facts for a list",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema39/100
Describe the returned fields so an agent can interpret the result.
{
"name": "quote_more_facts",
"title": "Price new facts for a list",
"description": "Prices reading new facts, such as a website or a founding year, for results the user already has in a list. The new facts are already in the run's reading, edited under a new ast_hash. The quote shows the most it can cost and when the quote expires. Nothing runs and nothing is spent.",
"inputSchema": {
"type": "object",
"title": "MoreFacts",
"required": [
"find_id",
"ast_hash"
],
"properties": {
"find_id": {
"type": "string",
"title": "Find Id",
"minLength": 1,
"description": "The find_id of the run whose results get the new facts."
},
"ast_hash": {
"type": "string",
"title": "Ast Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The ast_hash of the run's reading after the new facts were added to it."
},
"selected_entity_ids": {
"anyOf": [
{
"type": "array",
"items": {
"type": "string"
},
"maxItems": 10000,
"minItems": 1
},
{
"type": "null"
}
],
"title": "Selected Entity Ids",
"default": null,
"description": "The entity_id of each result to include. Without it, every result is included."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"scope": {
"type": "object"
},
"terms": {
"type": "object"
},
"currency": {
"type": "string"
},
"quote_id": {
"type": "string"
},
"whole_set": {
"type": "boolean"
},
"cost_basis": {
"type": "string"
},
"expires_at": {
"type": "string"
},
"quote_hash": {
"type": "string"
},
"can_authorize": {
"type": "boolean"
},
"maximum_cents": {
"type": "integer"
},
"pricing_version": {
"type": "string"
},
"existing_execution": {
"type": [
"object",
"null"
]
},
"retained_work_cents": {
"type": "integer"
},
"cost_profile_version": {
"type": "string"
}
}
},
"annotations": {
"title": "Price new facts for a list",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
Align the title with the capability actually described by this tool.
Description75/100
Reads the new facts for the list's results from a quote the user approved, and adds them to each result with their sources. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.
The description explains the capability, when to use it, and what to expect.
Inputs88/100
Clarify units, where to obtain this value, or conventions not already expressed in the schema.
quote_id
The quote_id of the quote the user approved.
Clarify units, where to obtain this value, or conventions not already expressed in the schema.
quote_hash
The quote_hash that came with that quote. It proves the quote is unchanged.
Clear enough to supply this argument.
confirm_max_cents
The most the user said yes to, in cents. It is at least the quote's maximum.
Clear enough to supply this argument.
idempotency_key
A key the caller makes once for this approval and sends again on a retry, so the work starts once.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Add new facts to a list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
Output schema26/100
Describe the returned fields so an agent can interpret the result.
{
"name": "run_more_facts",
"title": "Add new facts to a list",
"description": "Reads the new facts for the list's results from a quote the user approved, and adds them to each result with their sources. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.",
"inputSchema": {
"type": "object",
"title": "Run",
"required": [
"quote_id",
"quote_hash",
"confirm_max_cents",
"idempotency_key"
],
"properties": {
"quote_id": {
"type": "string",
"title": "Quote Id",
"minLength": 1,
"description": "The quote_id of the quote the user approved."
},
"quote_hash": {
"type": "string",
"title": "Quote Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The quote_hash that came with that quote. It proves the quote is unchanged."
},
"idempotency_key": {
"type": "string",
"title": "Idempotency Key",
"maxLength": 128,
"minLength": 8,
"description": "A key the caller makes once for this approval and sends again on a retry, so the work starts once."
},
"confirm_max_cents": {
"type": "integer",
"title": "Confirm Max Cents",
"minimum": 0,
"description": "The most the user said yes to, in cents. It is at least the quote's maximum."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"state": {
"type": "string"
},
"replay": {
"type": "boolean"
},
"find_id": {
"type": [
"string",
"null"
]
},
"maximum_cents": {
"type": "integer"
},
"reservation_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Add new facts to a list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description86/100
Prices checking a list's results again against the live web, to find what changed since the last run. The quote shows the most it can cost and when the quote expires. Nothing runs and nothing is spent.
The description explains the capability, when to use it, and what to expect.
Inputs86/100
Remove repeated schema validation from descriptions; focus on meaning and usage context.
find_id
The find_id of the run whose results are checked again.
Clear enough to supply this argument.
selected_entity_ids
The entity_id of each result to include. Without it, every result is included.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Price a recheck of a list",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema38/100
Describe the returned fields so an agent can interpret the result.
Name the specific action and subject instead of using a generic label such as ‘manage’ or ‘tool’.
Description80/100
Checks each result of the list again on the live web from a quote the user approved, and records what still holds and what changed. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.
Describe the returned information so the agent knows whether it will answer the request.
Inputs87/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
quote_id
The quote_id of the quote the user approved.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
quote_hash
The quote_hash that came with that quote. It proves the quote is unchanged.
Clear enough to supply this argument.
confirm_max_cents
The most the user said yes to, in cents. It is at least the quote's maximum.
Clear enough to supply this argument.
idempotency_key
A key the caller makes once for this approval and sends again on a retry, so the work starts once.
{
"name": "run_refresh",
"title": "Recheck a list",
"description": "Checks each result of the list again on the live web from a quote the user approved, and records what still holds and what changed. SpecStream reads the live web and charges for delivered results at the quoted rate, plus the checks the quote lists, never more than the approved amount. The same idempotency_key returns the same run instead of a second one. It refuses a quote priced for another job.",
"inputSchema": {
"type": "object",
"title": "Run",
"required": [
"quote_id",
"quote_hash",
"confirm_max_cents",
"idempotency_key"
],
"properties": {
"quote_id": {
"type": "string",
"title": "Quote Id",
"minLength": 1,
"description": "The quote_id of the quote the user approved."
},
"quote_hash": {
"type": "string",
"title": "Quote Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The quote_hash that came with that quote. It proves the quote is unchanged."
},
"idempotency_key": {
"type": "string",
"title": "Idempotency Key",
"maxLength": 128,
"minLength": 8,
"description": "A key the caller makes once for this approval and sends again on a retry, so the work starts once."
},
"confirm_max_cents": {
"type": "integer",
"title": "Confirm Max Cents",
"minimum": 0,
"description": "The most the user said yes to, in cents. It is at least the quote's maximum."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"state": {
"type": "string"
},
"replay": {
"type": "boolean"
},
"find_id": {
"type": [
"string",
"null"
]
},
"maximum_cents": {
"type": "integer"
},
"reservation_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Recheck a list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description96/100
Reads one quote by its quote_id: the most the work can cost in cents, what it covers, its terms, when it expires and whether it can still be approved, plus the run already started from it, if any. Nothing runs and nothing is spent.
The description explains the capability, when to use it, and what to expect.
Inputs71/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
quote_id
The quote_id that the price quote returned.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
Declare the result fields and their types in the output schema.
View output schema
{
"type": "object"
}
Original tool metadata
{
"name": "get_capacity_request",
"title": "Find work already started",
"description": "Reads the work started with an idempotency_key, so a retry that is not sure the first call worked does not start the work twice.",
"inputSchema": {
"type": "object",
"title": "McpGetCapacityRequestArgs",
"required": [
"request_key"
],
"properties": {
"project_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project's id, from the user's projects. Without it, the account's own work."
},
"request_key": {
"type": "string",
"title": "Request Key",
"minLength": 1,
"description": "The idempotency_key that started the work."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object"
},
"annotations": {
"title": "Find work already started",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description89/100
Reads the account's balance in cents: what is available, what is held for work in progress, and the monthly limit and what it has used. A balance is not permission to spend. Paid work starts only from a quote the user approved.
The description explains the capability, when to use it, and what to expect.
Inputs88/100
Input documentation gives an agent enough context to populate the arguments.
project_id
The project's id, from the user's projects. Without it, the account's own work.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
Fits when the user asks how a run is going or what it has cost so far. Not for reading its rows. Reads a run's request, its progress and its cost so far.
The description explains the capability, when to use it, and what to expect.
Inputs90/100
Input documentation gives an agent enough context to populate the arguments.
find_id
The run's id, as the start of the run or the run's progress returns it.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Check on a run",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema38/100
Describe the returned fields so an agent can interpret the result.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
Fits when a run has delivered results and the user wants the rows or the table. Not for a run's progress or cost. Reads the delivered results, one page at a time, with the source behind each condition a page states. A condition no page states yet is listed on its result. next_cursor is null on the last page. Every delivered result is charged. The markdown and csv formats return the asked columns in order as one table, with a source link for each value.
The description explains the capability, when to use it, and what to expect.
Inputs93/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
find_id
The run's id, as the start of the run or the run's progress returns it.
Clear enough to supply this argument.
cursor
The next_cursor of the previous page. Without it, the first page.
Clear enough to supply this argument.
limit
How many items one page holds, at most.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
evidence
short gives each value with its source, and cuts long values. full gives every source check and whole values.
Clear enough to supply this argument.
format
json (the default) returns each result with its values and sources. markdown and csv return one table of the asked columns, with a source link for each value.
{
"name": "list_find_rows",
"title": "Read a run's results",
"description": "Fits when a run has delivered results and the user wants the rows or the table. Not for a run's progress or cost. Reads the delivered results, one page at a time, with the source behind each condition a page states. A condition no page states yet is listed on its result. next_cursor is null on the last page. Every delivered result is charged. The markdown and csv formats return the asked columns in order as one table, with a source link for each value.",
"inputSchema": {
"type": "object",
"title": "McpListFindRowsArgs",
"required": [
"find_id"
],
"properties": {
"limit": {
"type": "integer",
"title": "Limit",
"default": 100,
"maximum": 100,
"minimum": 1,
"description": "How many items one page holds, at most."
},
"cursor": {
"anyOf": [
{
"type": "string",
"maxLength": 2048,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Cursor",
"default": null,
"description": "The next_cursor of the previous page. Without it, the first page."
},
"format": {
"enum": [
"json",
"markdown",
"csv"
],
"type": "string",
"title": "Format",
"default": "json",
"description": "json (the default) returns each result with its values and sources. markdown and csv return one table of the asked columns, with a source link for each value."
},
"find_id": {
"type": "string",
"title": "Find Id",
"description": "The run's id, as the start of the run or the run's progress returns it."
},
"evidence": {
"enum": [
"short",
"full"
],
"type": "string",
"title": "Evidence",
"default": "short",
"description": "short gives each value with its source, and cuts long values. full gives every source check and whole values."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"rows": {
"type": "array",
"items": {}
},
"table": {
"type": "string"
},
"format": {
"type": "string"
},
"status": {
"type": "string"
},
"columns": {
"type": "array",
"items": {}
},
"find_id": {
"type": [
"string",
"null"
]
},
"outcome": {
"type": [
"object",
"null"
]
},
"commerce": {
"type": [
"object",
"null"
]
},
"delivery": {
"type": [
"object",
"null"
]
},
"possible": {
"type": "array",
"items": {}
},
"flat_rows": {
"type": "array",
"items": {}
},
"entitlement": {
"type": "string"
},
"next_cursor": {
"type": [
"string",
"null"
]
},
"accepted_count": {
"type": "integer"
}
}
},
"annotations": {
"title": "Read a run's results",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Replays the saved condition checks of a run from their recorded evidence, and returns whether the replay gives the same results, with the number of checks replayed. No new search and no new charge.
The description explains the capability, when to use it, and what to expect.
Inputs84/100
Input documentation gives an agent enough context to populate the arguments.
find_id
The run's id, as the start of the run or the run's progress returns it.
{
"name": "replay_find_decisions",
"title": "Replay a run's checks",
"description": "Replays the saved condition checks of a run from their recorded evidence, and returns whether the replay gives the same results, with the number of checks replayed. No new search and no new charge.",
"inputSchema": {
"type": "object",
"title": "McpReplayFindDecisionsArgs",
"required": [
"find_id"
],
"properties": {
"find_id": {
"type": "string",
"title": "Find Id",
"minLength": 1,
"description": "The run's id, as the start of the run or the run's progress returns it."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"message": {
"type": "string"
},
"verified": {
"type": "boolean"
},
"available": {
"type": "boolean"
},
"decision_root": {
"type": [
"string",
"null"
]
},
"policy_version": {
"type": "string"
},
"decisions_checked": {
"type": "integer"
},
"qualifications_checked": {
"type": "integer"
}
}
},
"annotations": {
"title": "Replay a run's checks",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
{
"name": "list_your_fields",
"title": "Read a list's team fields",
"description": "Reads the status, checkbox and text fields a team added to a list, with the value of each field on each record and the fields' version.",
"inputSchema": {
"type": "object",
"title": "McpWorkspaceListYourFieldsArgs",
"required": [
"collection_id"
],
"properties": {
"collection_id": {
"type": "string",
"title": "Collection Id",
"minLength": 1,
"description": "The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"fields": {
"type": "array",
"items": {}
},
"values": {
"type": "object"
},
"version": {
"type": "integer"
},
"collection_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Read a list's team fields",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description89/100
Sets or clears the value of one team field, such as a status of Called, on the chosen records of a list, kept in the workspace database. Returns the fields and values with their new version. Nothing is charged.
The description explains the capability, when to use it, and what to expect.
Inputs92/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
expected_version
The version the latest read of it returned, 0 before the first save. The save fails with a conflict if someone changed it after that read.
Clear enough to supply this argument.
record_ids
The ids of the chosen records, from the list's records or the run's rows.
Clear enough to supply this argument.
key
The key of one team field on the list, from the list's team fields.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
value
The new value: text, or true or false for a check. Without it, the value clears.
Clear enough to supply this argument.
collection_id
The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Set a team field on records",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema29/100
Describe the returned fields so an agent can interpret the result.
{
"name": "set_your_field",
"title": "Set a team field on records",
"description": "Sets or clears the value of one team field, such as a status of Called, on the chosen records of a list, kept in the workspace database. Returns the fields and values with their new version. Nothing is charged.",
"inputSchema": {
"type": "object",
"title": "McpWorkspaceSetYourFieldValueArgs",
"required": [
"expected_version",
"record_ids",
"key",
"collection_id"
],
"properties": {
"key": {
"type": "string",
"title": "Key",
"maxLength": 96,
"minLength": 10,
"description": "The key of one team field on the list, from the list's team fields."
},
"value": {
"anyOf": [
{
"type": "string"
},
{
"type": "boolean"
},
{
"type": "null"
}
],
"title": "Value",
"default": null,
"description": "The new value: text, or true or false for a check. Without it, the value clears."
},
"record_ids": {
"type": "array",
"items": {
"type": "string"
},
"title": "Record Ids",
"maxItems": 2000,
"minItems": 1,
"description": "The ids of the chosen records, from the list's records or the run's rows."
},
"collection_id": {
"type": "string",
"title": "Collection Id",
"minLength": 1,
"description": "The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results."
},
"expected_version": {
"type": "integer",
"title": "Expected Version",
"minimum": 0,
"description": "The version the latest read of it returned, 0 before the first save. The save fails with a conflict if someone changed it after that read."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"fields": {
"type": "array",
"items": {}
},
"values": {
"type": "object"
},
"version": {
"type": "integer"
},
"collection_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Set a team field on records",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description91/100
Creates a link to a fixed copy of every record in a list, such as an imported list. Anyone with the link can open it without an account. Each value keeps the source link its list gives. Nothing is charged. Returns the share id, which turns the link off.
The description explains the capability, when to use it, and what to expect.
Inputs92/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
expires_days
Days the link stays on. null keeps it on.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
allow_export
Whether people with the link can download the rows.
Clear enough to supply this argument.
collection_id
The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results.
{
"name": "share_collection",
"title": "Share a list",
"description": "Creates a link to a fixed copy of every record in a list, such as an imported list. Anyone with the link can open it without an account. Each value keeps the source link its list gives. Nothing is charged. Returns the share id, which turns the link off.",
"inputSchema": {
"type": "object",
"title": "McpWorkspaceShareCollectionArgs",
"required": [
"collection_id"
],
"properties": {
"allow_export": {
"type": "boolean",
"title": "Allow Export",
"default": false,
"description": "Whether people with the link can download the rows."
},
"expires_days": {
"anyOf": [
{
"type": "integer",
"maximum": 365,
"minimum": 1
},
{
"type": "null"
}
],
"title": "Expires Days",
"default": 90,
"description": "Days the link stays on. null keeps it on."
},
"collection_id": {
"type": "string",
"title": "Collection Id",
"minLength": 1,
"description": "The list's id, from the user's lists or a workspace search. A run's find_id is also the id of its results."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"revoked": {
"type": "boolean"
},
"row_count": {
"type": "integer"
},
"created_at": {
"type": "string"
},
"expires_at": {
"type": [
"string",
"null"
]
},
"public_url": {
"type": "string"
},
"snapshot_id": {
"type": "string"
},
"allow_export": {
"type": "boolean"
}
}
},
"annotations": {
"title": "Share a list",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
}
{
"name": "revoke_share",
"title": "Stop sharing results",
"description": "Turns off a shared link to a run's results or to a list. The link stops working at once. The results and the list stay.",
"inputSchema": {
"type": "object",
"title": "McpRevokeShareArgs",
"required": [
"share_id"
],
"properties": {
"share_id": {
"type": "string",
"title": "Share Id",
"minLength": 1,
"description": "The share's id, as the share link returned it."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"revoked": {
"type": "boolean"
},
"share_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Stop sharing results",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description85/100
Fits when the user asks which lists they have. Not for finding one by name or text, which the workspace search does. Lists the imported lists and results the caller can open.
The description explains the capability, when to use it, and what to expect.
Inputs83/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
project_id
The project's id, from the user's projects. Without it, the account's own work.
Clear enough to supply this argument.
after
The next_after of the previous page. Without it, the first page.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
limit
How many items one page holds, at most.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description79/100
Fits when the user asks for something already in their SpecStream account, by name or text. Not for a search of the live web. Finds records, lists, requests, projects, people, notes and monitors by name or text.
The description explains the capability, when to use it, and what to expect.
Inputs83/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
q
Words to find in names and text.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
limit
How many items one page holds, at most.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
cursor
The next_cursor of the previous page. Without it, the first page.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
Fits when the user asks for one fact or a quick answer from the live web, or wants the first rows of a list before the full table is priced. Not for the text of known pages. Fast search: answers in seconds. Returns kind (fact or list), the answer with the page it was read from, rows with each result's name and the facts the request asks for, each value with its page, the search results read (sources), took_ms and request_id. num (1 to 100) sets how many sources come back and page (1 to 10) continues the same list; next_page names the next page. For a list, handoff names the estimated rows and the next steps, a compiled request and a quote for the full table. A call of up to 10 results takes $0.005 from the account's balance; each started block of 10 more results adds $0.006. With speed=fast a call takes $0.001 and a block $0.0012. It needs no quote or approval. A full list is run from a compiled request and its quote, not from this search.
The description explains the capability, when to use it, and what to expect.
Inputs90/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
q
The question or the request in everyday words, in any language. Search operators hold on every result: site:, -site:, -word, -"phrase", "phrase", filetype:, after: and before: (YYYY, YYYY-MM or YYYY-MM-DD). A result that breaks one is left out; when that leaves a page short, the engines' next pages are searched first.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
passages
Also return the text of the top pages closest to the request, each with its page, so a caller sees the evidence without reading a page
Clear enough to supply this argument.
num
The number of results (sources) to return, 1 to 100, or 1 to 200 with images=true. Without num or page, the reply holds every result the answer read. When the first search finds fewer, SpecStream searches deeper (more engines and their next result pages) before it answers. A call of up to 10 results is $0.005; each started block of 10 more results adds $0.006 (100 results are $0.059). On the fast tier (speed=fast) a call of up to 10 results is $0.001 and each started block of 10 more adds $0.0012 (100 results are $0.0118). An image search (images=true) always runs the fast tier, so 200 images are $0.0238. A call is charged for the results it returns; a call that returns none costs nothing.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
page
The page of results to return, 1 to 10, num results each (10 without num). Page 2 continues the same list as page 1. The answer and the rows are on page 1 only. next_page in the reply names the next page, or is null when the engines found no more.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
speed
fast gives the search results in about a second: the first engines' results with their snippets, each with the time it was read (read_at), with no answer, rows or passages; num and page past the first 10 results come from the engines' own later pages. default also reads the results for the answer and its rows. A fast call of up to 10 results is $0.001 ($1 per 1,000), a default call $0.005 ($5 per 1,000).
Clear enough to supply this argument.
news
News items for the query instead of web results: newest first, each with its publisher (source), its day (published) and its time (published_at). A fast search at the same price; not with passages.
Clear enough to supply this argument.
language
The language of the results, as a two-letter ISO 639-1 code (de). Sent to every search engine that takes a language.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
country
The country the results are for, as a two-letter ISO 3166-1 code (DE). Sent to every search engine that takes a country.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
safe
The safe search level: off, moderate or strict. Sent to every search engine that takes one; without it each engine uses its own default.
Clear enough to supply this argument.
engines
Narrows the engines this call may ask, never past the key's own engine_rule: engine names and two-letter ISO 3166-1 country codes of their operators, comma-separated. A leading - denies one (-CN asks no engine run from China); names without - are the only ones asked (bing,duckduckgo). The engines are listed at /v1/search-engines.
Clear enough to supply this argument.
engines_asked
Name the engines the query was sent to in the reply (engines_asked).
Clear enough to supply this argument.
pages_wait_ms
With passages=true: how long, in milliseconds, the call waits for the pages' text. The first five sources wait for theirs; the rest keep what is in by then, and pages_pending names the sources still being read. Absent: the call waits for every source's text, as before.
{
"name": "search_web",
"title": "Search the web",
"description": "Fits when the user asks for one fact or a quick answer from the live web, or wants the first rows of a list before the full table is priced. Not for the text of known pages. Fast search: answers in seconds. Returns kind (fact or list), the answer with the page it was read from, rows with each result's name and the facts the request asks for, each value with its page, the search results read (sources), took_ms and request_id. num (1 to 100) sets how many sources come back and page (1 to 10) continues the same list; next_page names the next page. For a list, handoff names the estimated rows and the next steps, a compiled request and a quote for the full table. A call of up to 10 results takes $0.005 from the account's balance; each started block of 10 more results adds $0.006. With speed=fast a call takes $0.001 and a block $0.0012. It needs no quote or approval. A full list is run from a compiled request and its quote, not from this search.",
"inputSchema": {
"type": "object",
"title": "McpWebSearchArgs",
"required": [
"q"
],
"properties": {
"q": {
"type": "string",
"title": "Q",
"maxLength": 2000,
"minLength": 1,
"description": "The question or the request in everyday words, in any language. Search operators hold on every result: site:, -site:, -word, -\"phrase\", \"phrase\", filetype:, after: and before: (YYYY, YYYY-MM or YYYY-MM-DD). A result that breaks one is left out; when that leaves a page short, the engines' next pages are searched first."
},
"num": {
"anyOf": [
{
"type": "integer",
"maximum": 200,
"minimum": 1
},
{
"type": "null"
}
],
"title": "Num",
"default": null,
"description": "The number of results (sources) to return, 1 to 100, or 1 to 200 with images=true. Without num or page, the reply holds every result the answer read. When the first search finds fewer, SpecStream searches deeper (more engines and their next result pages) before it answers. A call of up to 10 results is $0.005; each started block of 10 more results adds $0.006 (100 results are $0.059). On the fast tier (speed=fast) a call of up to 10 results is $0.001 and each started block of 10 more adds $0.0012 (100 results are $0.0118). An image search (images=true) always runs the fast tier, so 200 images are $0.0238. A call is charged for the results it returns; a call that returns none costs nothing."
},
"news": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"title": "News",
"default": null,
"description": "News items for the query instead of web results: newest first, each with its publisher (source), its day (published) and its time (published_at). A fast search at the same price; not with passages."
},
"page": {
"anyOf": [
{
"type": "integer",
"maximum": 10,
"minimum": 1
},
{
"type": "null"
}
],
"title": "Page",
"default": null,
"description": "The page of results to return, 1 to 10, num results each (10 without num). Page 2 continues the same list as page 1. The answer and the rows are on page 1 only. next_page in the reply names the next page, or is null when the engines found no more."
},
"safe": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Safe",
"default": null,
"description": "The safe search level: off, moderate or strict. Sent to every search engine that takes one; without it each engine uses its own default."
},
"speed": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Speed",
"default": null,
"description": "fast gives the search results in about a second: the first engines' results with their snippets, each with the time it was read (read_at), with no answer, rows or passages; num and page past the first 10 results come from the engines' own later pages. default also reads the results for the answer and its rows. A fast call of up to 10 results is $0.001 ($1 per 1,000), a default call $0.005 ($5 per 1,000)."
},
"country": {
"anyOf": [
{
"type": "string",
"pattern": "^[A-Za-z]{2}$"
},
{
"type": "null"
}
],
"title": "Country",
"default": null,
"description": "The country the results are for, as a two-letter ISO 3166-1 code (DE). Sent to every search engine that takes a country."
},
"engines": {
"anyOf": [
{
"type": "string",
"maxLength": 2000
},
{
"type": "null"
}
],
"title": "Engines",
"default": null,
"description": "Narrows the engines this call may ask, never past the key's own engine_rule: engine names and two-letter ISO 3166-1 country codes of their operators, comma-separated. A leading - denies one (-CN asks no engine run from China); names without - are the only ones asked (bing,duckduckgo). The engines are listed at /v1/search-engines."
},
"language": {
"anyOf": [
{
"type": "string",
"pattern": "^[A-Za-z]{2}$"
},
{
"type": "null"
}
],
"title": "Language",
"default": null,
"description": "The language of the results, as a two-letter ISO 639-1 code (de). Sent to every search engine that takes a language."
},
"passages": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"title": "Passages",
"default": null,
"description": "Also return the text of the top pages closest to the request, each with its page, so a caller sees the evidence without reading a page"
},
"engines_asked": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"title": "Engines Asked",
"default": null,
"description": "Name the engines the query was sent to in the reply (engines_asked)."
},
"pages_wait_ms": {
"anyOf": [
{
"type": "integer",
"maximum": 3500,
"minimum": 0
},
{
"type": "null"
}
],
"title": "Pages Wait Ms",
"default": null,
"description": "With passages=true: how long, in milliseconds, the call waits for the pages' text. The first five sources wait for theirs; the rest keep what is in by then, and pages_pending names the sources still being read. Absent: the call waits for every source's text, as before."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"id": {
"type": [
"string",
"null"
]
},
"kind": {
"type": "string"
},
"news": {
"type": [
"boolean",
"null"
]
},
"page": {
"type": [
"integer",
"null"
]
},
"rows": {
"type": "array",
"items": {}
},
"query": {
"type": "string"
},
"answer": {
"type": [
"object",
"null"
]
},
"images": {
"type": [
"boolean",
"null"
]
},
"columns": {
"type": "array",
"items": {}
},
"handoff": {
"type": [
"object",
"null"
]
},
"read_at": {
"type": [
"string",
"null"
]
},
"sources": {
"type": "array",
"items": {}
},
"took_ms": {
"type": "integer"
},
"passages": {
"type": [
"array",
"null"
],
"items": {}
},
"next_page": {
"type": [
"integer",
"null"
]
},
"operators": {
"type": [
"array",
"null"
],
"items": {}
},
"price_usd": {
"type": "number"
},
"request_id": {
"type": "string"
},
"engines_asked": {
"type": [
"array",
"null"
],
"items": {}
},
"pages_pending": {
"type": [
"array",
"null"
],
"items": {}
},
"corrected_query": {
"type": [
"string",
"null"
]
}
}
},
"annotations": {
"title": "Search the web",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description98/100
Fits when the user gives URLs, or wants the full text of pages a search returned. Not for finding pages or answering a question, which the fast web search does, or for building a list. Reads up to 20 web pages or documents (PDF, Word, Excel) live and returns each one's main text, as Markdown (format markdown, the default) or plain text (format text), with its title and the date it says it was published. Each result has status: ok, blocked, not_found, empty, unsupported, invalid_url or failed, and read_from: live, or site or archive when the page refused a live read and its site's feed or its newest archive capture gave the text (read_from_url names the feed, captured_at says when the capture was taken). A page is charged only when its status is ok, $0.0008 a page from the account's balance, with no quote or approval.
The description explains the capability, when to use it, and what to expect.
Inputs70/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
urls
The addresses of the pages, 20 at most.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › ExtractHtml › html
The page's HTML.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
$defs › ExtractHtml › url
The page's address, for its links and source.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
format
markdown (the default), text or html.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
include
Extra parts in the result: html, links, response or full_text.
Explain the meaning of this value, not just its name.
{
"name": "read_pages",
"title": "Read pages",
"description": "Fits when the user gives URLs, or wants the full text of pages a search returned. Not for finding pages or answering a question, which the fast web search does, or for building a list. Reads up to 20 web pages or documents (PDF, Word, Excel) live and returns each one's main text, as Markdown (format markdown, the default) or plain text (format text), with its title and the date it says it was published. Each result has status: ok, blocked, not_found, empty, unsupported, invalid_url or failed, and read_from: live, or site or archive when the page refused a live read and its site's feed or its newest archive capture gave the text (read_from_url names the feed, captured_at says when the capture was taken). A page is charged only when its status is ok, $0.0008 a page from the account's balance, with no quote or approval.",
"inputSchema": {
"type": "object",
"$defs": {
"ExtractHtml": {
"type": "object",
"title": "ExtractHtml",
"required": [
"html"
],
"properties": {
"url": {
"anyOf": [
{
"type": "string",
"maxLength": 2048
},
{
"type": "null"
}
],
"title": "Url",
"default": null,
"description": "The page's address, for its links and source."
},
"html": {
"type": "string",
"title": "Html",
"maxLength": 8388608,
"minLength": 1,
"description": "The page's HTML."
}
},
"description": "A page already held by the caller, sent as its HTML. Nothing is fetched.",
"additionalProperties": false
}
},
"title": "McpExtractPagesArgs",
"properties": {
"urls": {
"type": "array",
"items": {
"type": "string"
},
"title": "Urls",
"maxItems": 20,
"description": "The addresses of the pages, 20 at most."
},
"pages": {
"type": "array",
"items": {
"$ref": "#/$defs/ExtractHtml"
},
"title": "Pages",
"maxItems": 20,
"description": "Page HTML the caller already has, read without a fetch."
},
"format": {
"enum": [
"markdown",
"text",
"html"
],
"type": "string",
"title": "Format",
"default": "markdown",
"description": "markdown (the default), text or html."
},
"include": {
"type": "array",
"items": {
"enum": [
"html",
"links",
"response",
"full_text"
],
"type": "string"
},
"title": "Include",
"description": "Extra parts in the result: html, links, response or full_text."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"results": {
"type": "array",
"items": {}
},
"took_ms": {
"type": "integer"
},
"price_usd": {
"type": "number"
},
"request_id": {
"type": "string"
}
}
},
"annotations": {
"title": "Read pages",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
Lists the account's past fast searches, newest first: each question with when it was asked, its kind (fact or list), the answer and the page it was read from, the number of rows, and the search results its values come from (sources). q keeps the searches whose words contain it. An answer the account already paid for is read here again without a new search. next_cursor leads to older searches. Nothing is charged.
The description explains the capability, when to use it, and what to expect.
Inputs83/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
q
Searches whose words contain this text, in any case
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
limit
How many items one page holds, at most.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
cursor
The next_cursor of the previous page. Without it, the newest.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "List your web searches",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema28/100
Describe the returned fields so an agent can interpret the result.
{
"name": "list_web_searches",
"title": "List your web searches",
"description": "Lists the account's past fast searches, newest first: each question with when it was asked, its kind (fact or list), the answer and the page it was read from, the number of rows, and the search results its values come from (sources). q keeps the searches whose words contain it. An answer the account already paid for is read here again without a new search. next_cursor leads to older searches. Nothing is charged.",
"inputSchema": {
"type": "object",
"title": "McpWebSearchAnswersArgs",
"properties": {
"q": {
"anyOf": [
{
"type": "string",
"maxLength": 2000,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Q",
"default": null,
"description": "Searches whose words contain this text, in any case"
},
"limit": {
"anyOf": [
{
"type": "integer",
"maximum": 50,
"minimum": 1
},
{
"type": "null"
}
],
"title": "Limit",
"default": null,
"description": "How many items one page holds, at most."
},
"cursor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Cursor",
"default": null,
"description": "The next_cursor of the previous page. Without it, the newest."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"results": {
"type": "array",
"items": {}
},
"next_cursor": {
"type": [
"string",
"null"
]
}
}
},
"annotations": {
"title": "List your web searches",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description87/100
Opens one past fast search by its id, from the account's search history or from the answer of the search itself, exactly as it was given: the answer, the rows with the page of each value, the search results read and, for a list, the hand-off to the full table. Nothing is charged.
The description explains the capability, when to use it, and what to expect.
Inputs72/100
Clarify ID sources, units, or usage conventions not already expressed in the schema.
answer_id
The search's id: a fast search's request_id, or an id from the search history.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Read a past web search",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema36/100
Describe the returned fields so an agent can interpret the result.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Fits when the user asks for a written report on a question, before it is approved. Returns the price of one report at each depth (quick, standard and deep) in dollars and in cents, and how many parts of the question each depth researches. A report is charged once, when it is delivered; a run that fails, is cancelled or reads no page is not charged. Nothing runs and nothing is charged.
The description explains the capability, when to use it, and what to expect.
Inputs75/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
depth
quick, standard or deep. Without it, the price of each depth.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
Annotations100/100
All four behavior hints are declared.
{
"title": "Price a research report",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema31/100
Declare the result fields and their types in the output schema.
View output schema
{
"type": "object",
"properties": {
"prices": {
"type": "array",
"items": {},
"description": "One price for each depth."
},
"charged": {
"type": "string",
"description": "When the price is charged."
}
},
"description": "The price of one report at each depth, charged once on delivery."
}
Original tool metadata
{
"name": "research_price",
"title": "Price a research report",
"description": "Fits when the user asks for a written report on a question, before it is approved. Returns the price of one report at each depth (quick, standard and deep) in dollars and in cents, and how many parts of the question each depth researches. A report is charged once, when it is delivered; a run that fails, is cancelled or reads no page is not charged. Nothing runs and nothing is charged.",
"inputSchema": {
"type": "object",
"title": "ResearchPriceArgs",
"properties": {
"depth": {
"anyOf": [
{
"enum": [
"quick",
"standard",
"deep"
],
"type": "string"
},
{
"type": "null"
}
],
"title": "Depth",
"default": null,
"description": "quick, standard or deep. Without it, the price of each depth."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"prices": {
"type": "array",
"items": {},
"description": "One price for each depth."
},
"charged": {
"type": "string",
"description": "When the price is charged."
}
},
"description": "The price of one report at each depth, charged once on delivery."
},
"annotations": {
"title": "Price a research report",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description98/100
Fits when the user wants a written report that answers a question from the live web, with the page behind each cited sentence, and has approved its price. Not for one fact or a quick answer, which the fast web search gives, or for a table of results. Starts a run that splits the question into parts, searches the live web, reads pages and writes a markdown report with numbered citations, or JSON that matches a given JSON Schema. A quick report takes about a minute, a deep one up to about fifteen. Returns the run's id, its status and its price at once. The report is charged once, at delivery, at its depth's price ($0.10 quick, $0.30 standard, $1.00 deep) from the account's balance; a failed or cancelled run is not charged. A depth priced above confirm_max_cents is refused, and the same idempotency_key returns the same run.
The description explains the capability, when to use it, and what to expect.
Inputs86/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
question
The question in plain words, in any language.
Clear enough to supply this argument.
depth
quick, standard or deep: how far the run searches and reads, and its price.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
confirm_max_cents
The most the user approved for this report, in cents. A depth priced above it is refused and nothing runs.
Clear enough to supply this argument.
idempotency_key
A key that stays the same if this call is retried, so the run starts once.
Clear enough to supply this argument.
output_schema
A JSON Schema for the answer. Without it, the report is markdown text.
Clear enough to supply this argument.
length
short, standard or long: the target length of the text, not of the research.
Clear enough to supply this argument.
language
The report's language. Without it, the language of the question.
Clear enough to supply this argument.
include_domains
Sites the run reads from, and no others.
Clear enough to supply this argument.
exclude_domains
Sites the run never reads.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Write a research report",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
Output schema45/100
Describe the returned fields so an agent can interpret the result.
{
"name": "start_research",
"title": "Write a research report",
"description": "Fits when the user wants a written report that answers a question from the live web, with the page behind each cited sentence, and has approved its price. Not for one fact or a quick answer, which the fast web search gives, or for a table of results. Starts a run that splits the question into parts, searches the live web, reads pages and writes a markdown report with numbered citations, or JSON that matches a given JSON Schema. A quick report takes about a minute, a deep one up to about fifteen. Returns the run's id, its status and its price at once. The report is charged once, at delivery, at its depth's price ($0.10 quick, $0.30 standard, $1.00 deep) from the account's balance; a failed or cancelled run is not charged. A depth priced above confirm_max_cents is refused, and the same idempotency_key returns the same run.",
"inputSchema": {
"type": "object",
"title": "ResearchArgs",
"required": [
"question",
"confirm_max_cents",
"idempotency_key"
],
"properties": {
"depth": {
"enum": [
"quick",
"standard",
"deep"
],
"type": "string",
"title": "Depth",
"default": "standard",
"description": "quick, standard or deep: how far the run searches and reads, and its price."
},
"length": {
"enum": [
"short",
"standard",
"long"
],
"type": "string",
"title": "Length",
"default": "standard",
"description": "short, standard or long: the target length of the text, not of the research."
},
"language": {
"anyOf": [
{
"type": "string",
"maxLength": 35,
"minLength": 2
},
{
"type": "null"
}
],
"title": "Language",
"default": null,
"description": "The report's language. Without it, the language of the question."
},
"question": {
"type": "string",
"title": "Question",
"maxLength": 15000,
"minLength": 1,
"description": "The question in plain words, in any language."
},
"output_schema": {
"anyOf": [
{
"type": "object",
"additionalProperties": true
},
{
"type": "null"
}
],
"title": "Output Schema",
"default": null,
"description": "A JSON Schema for the answer. Without it, the report is markdown text."
},
"exclude_domains": {
"anyOf": [
{
"type": "array",
"items": {
"type": "string"
},
"maxItems": 100
},
{
"type": "null"
}
],
"title": "Exclude Domains",
"default": null,
"description": "Sites the run never reads."
},
"idempotency_key": {
"type": "string",
"title": "Idempotency Key",
"maxLength": 128,
"minLength": 8,
"description": "A key that stays the same if this call is retried, so the run starts once."
},
"include_domains": {
"anyOf": [
{
"type": "array",
"items": {
"type": "string"
},
"maxItems": 100
},
{
"type": "null"
}
],
"title": "Include Domains",
"default": null,
"description": "Sites the run reads from, and no others."
},
"confirm_max_cents": {
"type": "integer",
"title": "Confirm Max Cents",
"minimum": 0,
"description": "The most the user approved for this report, in cents. A depth priced above it is refused and nothing runs."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"depth": {
"type": "string"
},
"links": {
"type": "object"
},
"status": {
"type": "string"
},
"price_usd": {
"type": "number"
},
"created_at": {
"type": "string"
}
}
},
"annotations": {
"title": "Write a research report",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description94/100
Reads a research run: its status, its stage (planning, searching, reading, writing, checking or done), how many searches, pages and sources it has, and the report once the run is completed. The report is markdown with numbered citations, or JSON for a JSON Schema with the citations of each field. Each citation names the page, its title, when it was read, the passage and whether that passage supports the sentence. A call with a progress token gets progress notifications while the run works, for up to 45 seconds. Nothing is charged.
The description explains the capability, when to use it, and what to expect.
Inputs75/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
id
The run's id, as the run's start gave it.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
Annotations100/100
All four behavior hints are declared.
{
"title": "Read a research report",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema37/100
Describe the returned fields so an agent can interpret the result.
{
"name": "get_research",
"title": "Read a research report",
"description": "Reads a research run: its status, its stage (planning, searching, reading, writing, checking or done), how many searches, pages and sources it has, and the report once the run is completed. The report is markdown with numbered citations, or JSON for a JSON Schema with the citations of each field. Each citation names the page, its title, when it was read, the passage and whether that passage supports the sentence. A call with a progress token gets progress notifications while the run works, for up to 45 seconds. Nothing is charged.",
"inputSchema": {
"type": "object",
"title": "ResearchRunArgs",
"required": [
"id"
],
"properties": {
"id": {
"type": "string",
"title": "Id",
"minLength": 1,
"description": "The run's id, as the run's start gave it."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"depth": {
"type": "string"
},
"error": {
"type": [
"object",
"null"
]
},
"input": {
"type": "string"
},
"stage": {
"type": "string"
},
"usage": {
"type": "object"
},
"report": {
"type": [
"object",
"null"
]
},
"status": {
"type": "string"
},
"metadata": {
"type": "object"
},
"progress": {
"type": "object"
},
"created_at": {
"type": "string"
},
"started_at": {
"type": [
"string",
"null"
]
},
"completed_at": {
"type": [
"string",
"null"
]
}
}
},
"annotations": {
"title": "Read a research report",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description90/100
Lists the account's research runs, newest first, each with its status, depth and progress; next_after continues the list. Nothing is changed or charged.
The description explains the capability, when to use it, and what to expect.
Inputs62/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
limit
Runs on one page.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
after
The next_after of the page before.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
{
"name": "get_spec_revision",
"title": "Read one version of a saved request",
"description": "Reads one version of a saved request as it was saved: its reading of the request with the ast_hash, its name and notes, and the version it came from.",
"inputSchema": {
"type": "object",
"title": "McpGetSpecRevisionArgs",
"required": [
"spec_id",
"revision_id"
],
"properties": {
"spec_id": {
"type": "string",
"title": "Spec Id",
"minLength": 1,
"description": "The spec_id of a saved request, from the saved requests or a workspace search."
},
"revision_id": {
"type": "string",
"title": "Revision Id",
"minLength": 1,
"description": "The revision_id of one version of the saved request, from its version history. The current version's id is the request's head_revision_id."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"ast": {},
"name": {
"type": "string"
},
"label": {
"type": [
"string",
"null"
]
},
"context": {
"type": "string"
},
"spec_id": {
"type": "string"
},
"ast_hash": {
"type": "string"
},
"revision": {
"type": "integer"
},
"created_at": {
"type": "string"
},
"revision_id": {
"type": "string"
},
"parent_revision_id": {
"type": [
"string",
"null"
]
}
}
},
"annotations": {
"title": "Read one version of a saved request",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description86/100
Reads one project: its name, its owner, the caller's role in it, whether the caller can spend in it, its monthly spending limit in cents and its version.
The description explains the capability, when to use it, and what to expect.
Inputs77/100
Input documentation gives an agent enough context to populate the arguments.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
Fits when the user already has a list as a CSV or Excel file and wants details filled in or each row checked. Not for finding new rows on the web, or for a linked Google Sheet or Airtable view. Imports the file with its names, websites, places and any other columns. Nothing is researched or charged. Returns the import id, which a quote takes as source_import_id to fill in details or check each row.
The description explains the capability, when to use it, and what to expect.
Inputs78/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
import_id
An id the caller chooses for the new list: 8 to 128 letters, digits, - or _. A retry with the same id does not import twice.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
name
The name of the new list.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
format
csv or xlsx.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
project_id
The project's id, from the user's projects. Without it, the account's own work.
Clear enough to supply this argument.
sheet
For xlsx, the sheet to read. Without it, the first sheet.
Clear enough to supply this argument.
content
The CSV file as text. Each import takes content or content_base64.
Resolve the mismatch between this description and its schema constraints.
content_base64
The file as base64. An xlsx file needs it.
Clear enough to supply this argument.
mapping
Which source column fills each field, as {field: column header}. A field not named here is matched to a column by its header.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Import a CSV or Excel file",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema43/100
Describe the returned fields so an agent can interpret the result.
{
"name": "import_csv",
"title": "Import a CSV or Excel file",
"description": "Fits when the user already has a list as a CSV or Excel file and wants details filled in or each row checked. Not for finding new rows on the web, or for a linked Google Sheet or Airtable view. Imports the file with its names, websites, places and any other columns. Nothing is researched or charged. Returns the import id, which a quote takes as source_import_id to fill in details or check each row.",
"inputSchema": {
"type": "object",
"title": "McpImportFileCommandArgs",
"required": [
"import_id",
"name",
"format"
],
"properties": {
"name": {
"type": "string",
"title": "Name",
"maxLength": 120,
"minLength": 1,
"description": "The name of the new list."
},
"sheet": {
"anyOf": [
{
"type": "string",
"maxLength": 31
},
{
"type": "null"
}
],
"title": "Sheet",
"default": null,
"description": "For xlsx, the sheet to read. Without it, the first sheet."
},
"format": {
"type": "string",
"title": "Format",
"pattern": "^(csv|xlsx)$",
"description": "csv or xlsx."
},
"content": {
"anyOf": [
{
"type": "string",
"maxLength": 20000000
},
{
"type": "null"
}
],
"title": "Content",
"default": null,
"description": "The CSV file as text. Each import takes content or content_base64."
},
"mapping": {
"anyOf": [
{
"type": "object",
"maxProperties": 100,
"additionalProperties": {
"type": "string"
}
},
{
"type": "null"
}
],
"title": "Mapping",
"default": null,
"description": "Which source column fills each field, as {field: column header}. A field not named here is matched to a column by its header."
},
"import_id": {
"type": "string",
"title": "Import Id",
"pattern": "^[A-Za-z0-9_-]+$",
"maxLength": 128,
"minLength": 8,
"description": "An id the caller chooses for the new list: 8 to 128 letters, digits, - or _. A retry with the same id does not import twice."
},
"project_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project's id, from the user's projects. Without it, the account's own work."
},
"content_base64": {
"anyOf": [
{
"type": "string",
"maxLength": 26666668
},
{
"type": "null"
}
],
"title": "Content Base64",
"default": null,
"description": "The file as base64. An xlsx file needs it."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"rows": {
"type": "array",
"items": {}
},
"imported": {
"type": "boolean"
},
"max_rows": {
"type": "integer"
},
"metadata": {
"type": "object"
},
"account_id": {
"type": "string"
},
"created_at": {
"type": "string"
},
"project_id": {
"type": [
"string",
"null"
]
},
"total_rows": {
"type": "integer"
},
"rejected_rows": {
"type": "array",
"items": {}
},
"imported_count": {
"type": "integer"
}
}
},
"annotations": {
"title": "Import a CSV or Excel file",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
Fits when the user names one public page that holds records and can copy 3 to 8 example rows from it. Not for the text of a page or for a list from the whole web. Reads the page and learns its records from the example rows. Returns a preview and its explanation: how each field is read, which rows are the user's examples and which were inferred, and what happens when the page changes.
The description explains the capability, when to use it, and what to expect.
Inputs89/100
Explain the meaning of this value, not just its name.
url
The public page that holds the records.
Clear enough to supply this argument.
name
The source's name.
Explain the meaning of this value, not just its name.
name_column
The column of the examples that names each record.
Clear enough to supply this argument.
examples
3 to 8 rows the user copied from the page, as {column: value}.
Clear enough to supply this argument.
exclude
Text of page parts that are not records, such as a footer.
Clear enough to supply this argument.
project_id
The project's id, from the user's projects. Without it, the account's own work.
{
"name": "teach_web_source",
"title": "Teach a website source",
"description": "Fits when the user names one public page that holds records and can copy 3 to 8 example rows from it. Not for the text of a page or for a list from the whole web. Reads the page and learns its records from the example rows. Returns a preview and its explanation: how each field is read, which rows are the user's examples and which were inferred, and what happens when the page changes.",
"inputSchema": {
"type": "object",
"title": "McpTeachWebSourceArgs",
"required": [
"url",
"name",
"examples"
],
"properties": {
"url": {
"type": "string",
"title": "Url",
"maxLength": 4096,
"minLength": 8,
"description": "The public page that holds the records."
},
"name": {
"type": "string",
"title": "Name",
"maxLength": 120,
"minLength": 1,
"description": "The source's name."
},
"exclude": {
"type": "array",
"items": {
"type": "string"
},
"title": "Exclude",
"maxItems": 16,
"description": "Text of page parts that are not records, such as a footer."
},
"examples": {
"type": "array",
"items": {
"type": "object",
"additionalProperties": {
"type": "string"
}
},
"title": "Examples",
"maxItems": 8,
"minItems": 3,
"description": "3 to 8 rows the user copied from the page, as {column: value}."
},
"project_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project's id, from the user's projects. Without it, the account's own work."
},
"name_column": {
"type": "string",
"title": "Name Column",
"default": "name",
"maxLength": 80,
"minLength": 1,
"description": "The column of the examples that names each record."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"rows": {
"type": "array",
"items": {}
},
"count": {
"type": "integer"
},
"reuse": {
"type": "string"
},
"notice": {
"type": "string"
},
"headers": {
"type": "array",
"items": {}
},
"source_id": {
"type": "string"
},
"fetched_at": {
"type": "string"
},
"preview_id": {
"type": "string"
},
"source_url": {
"type": "string"
},
"explanation": {
"type": [
"object",
"null"
]
},
"name_column": {
"type": "string"
},
"preview_hash": {
"type": "string"
},
"missing_cells": {
"type": "integer"
},
"source_sha256": {
"type": "string"
},
"qualification_authority": {
"type": "boolean"
}
}
},
"annotations": {
"title": "Teach a website source",
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description79/100
Saves a preview the user reviewed as an imported list, by its preview hash. The user sees the preview and its explanation before the save. Saving does not research or spend.
The description explains the capability, when to use it, and what to expect.
Inputs83/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
preview_id
The preview's id, as the source preview returned it.
Clear enough to supply this argument.
preview_hash
The preview_hash of the preview the user reviewed, unchanged.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
import_id
An id the caller chooses for the new list: 8 to 128 letters, digits, - or _. A retry with the same id does not import twice.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
name
The name of the new list.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
mapping
Which source column fills each field, as {field: column header}. A field not named here is matched to a column by its header.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Save a source preview as a list",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema43/100
Describe the returned fields so an agent can interpret the result.
{
"name": "save_web_source_list",
"title": "Save a source preview as a list",
"description": "Saves a preview the user reviewed as an imported list, by its preview hash. The user sees the preview and its explanation before the save. Saving does not research or spend.",
"inputSchema": {
"type": "object",
"title": "McpSaveWebSourceListArgs",
"required": [
"preview_id",
"preview_hash",
"import_id",
"name"
],
"properties": {
"name": {
"type": "string",
"title": "Name",
"maxLength": 120,
"minLength": 1,
"description": "The name of the new list."
},
"mapping": {
"type": "object",
"title": "Mapping",
"description": "Which source column fills each field, as {field: column header}. A field not named here is matched to a column by its header.",
"additionalProperties": {
"type": "string"
}
},
"import_id": {
"type": "string",
"title": "Import Id",
"pattern": "^[A-Za-z0-9_-]+$",
"maxLength": 128,
"minLength": 8,
"description": "An id the caller chooses for the new list: 8 to 128 letters, digits, - or _. A retry with the same id does not import twice."
},
"preview_id": {
"type": "string",
"title": "Preview Id",
"maxLength": 128,
"minLength": 8,
"description": "The preview's id, as the source preview returned it."
},
"preview_hash": {
"type": "string",
"title": "Preview Hash",
"pattern": "^[0-9a-f]{64}$",
"description": "The preview_hash of the preview the user reviewed, unchanged."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"rows": {
"type": "array",
"items": {}
},
"imported": {
"type": "boolean"
},
"max_rows": {
"type": "integer"
},
"metadata": {
"type": "object"
},
"account_id": {
"type": "string"
},
"created_at": {
"type": "string"
},
"project_id": {
"type": [
"string",
"null"
]
},
"total_rows": {
"type": "integer"
},
"rejected_rows": {
"type": "array",
"items": {}
},
"imported_count": {
"type": "integer"
}
}
},
"annotations": {
"title": "Save a source preview as a list",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Opens the SpecStream view in the conversation, with the person's saved lists and the request editor. Returns the lists and the list settings. Nothing is changed or charged.
The description explains the capability, when to use it, and what to expect.
Inputs100/100
This tool declares no inputs, so argument descriptions are not needed.
Expand internal abbreviations and remove technical prefixes from the display title.
Description93/100
Shows the saved lists the person can open as a compact selector card in the conversation. Returns the lists and the list settings. Nothing is changed or charged.
The description explains the capability, when to use it, and what to expect.
Inputs100/100
This tool declares no inputs, so argument descriptions are not needed.
Describe the returned fields so an agent can interpret the result.
View output schema
{
"type": "object",
"properties": {
"items": {
"type": "array",
"items": {},
"description": "The saved lists."
},
"surface": {
"type": "string",
"description": "lists or panel: where the lists show."
},
"next_cursor": {
"type": [
"string",
"null"
]
},
"preferences": {
"type": "object",
"description": "The account's list layout and order."
}
},
"description": "The person's saved lists, with the list settings."
}
Original tool metadata
{
"name": "show_specstream_lists",
"title": "Show saved lists",
"description": "Shows the saved lists the person can open as a compact selector card in the conversation. Returns the lists and the list settings. Nothing is changed or charged.",
"inputSchema": {
"type": "object",
"required": [],
"properties": {},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"items": {
"type": "array",
"items": {},
"description": "The saved lists."
},
"surface": {
"type": "string",
"description": "lists or panel: where the lists show."
},
"next_cursor": {
"type": [
"string",
"null"
]
},
"preferences": {
"type": "object",
"description": "The account's list layout and order."
}
},
"description": "The person's saved lists, with the list settings."
},
"annotations": {
"title": "Show saved lists",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
Expand internal abbreviations and remove technical prefixes from the display title.
Description96/100
Shows a list and its rows as a card in the conversation. It takes a run's find_id or a saved list's id as collection_id, and returns the list with its columns and rows. Nothing is changed or charged.
The description explains the capability, when to use it, and what to expect.
Inputs75/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
collection_id
A run's find_id or a saved list's id.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
Describe the returned fields so an agent can interpret the result.
View output schema
{
"type": "object",
"properties": {
"surface": {
"type": "string",
"description": "records."
},
"collection": {
"type": "object",
"description": "The list, with its columns and rows."
},
"collection_id": {
"type": "string",
"description": "The id of the list shown."
}
},
"description": "One list with its rows."
}
Original tool metadata
{
"name": "show_specstream_records",
"title": "Show a list's results",
"description": "Shows a list and its rows as a card in the conversation. It takes a run's find_id or a saved list's id as collection_id, and returns the list with its columns and rows. Nothing is changed or charged.",
"inputSchema": {
"type": "object",
"required": [
"collection_id"
],
"properties": {
"collection_id": {
"type": "string",
"maxLength": 300,
"minLength": 1,
"description": "A run's find_id or a saved list's id."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"surface": {
"type": "string",
"description": "records."
},
"collection": {
"type": "object",
"description": "The list, with its columns and rows."
},
"collection_id": {
"type": "string",
"description": "The id of the list shown."
}
},
"description": "One list with its rows."
},
"annotations": {
"title": "Show a list's results",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
Shows a list file the person opened (.csv, .tsv, .xlsx, .geojson or .kml) as a list in the conversation. Returns the file's name and resource URI. The file is never changed, and nothing is imported or charged.
The description explains the capability, when to use it, and what to expect.
Inputs75/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
file › name
The file name.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
file › resourceUri
The file's resource URI from the host.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
Annotations100/100
All four behavior hints are declared.
{
"title": "Open a list file",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema53/100
Describe the returned fields so an agent can interpret the result.
View output schema
{
"type": "object",
"properties": {
"file": {
"type": "object",
"description": "The file's name and resourceUri."
},
"surface": {
"type": "string",
"description": "file."
}
},
"description": "The file the person opened, as a list."
}
Original tool metadata
{
"name": "open_list_file",
"title": "Open a list file",
"description": "Shows a list file the person opened (.csv, .tsv, .xlsx, .geojson or .kml) as a list in the conversation. Returns the file's name and resource URI. The file is never changed, and nothing is imported or charged.",
"inputSchema": {
"type": "object",
"required": [
"file"
],
"properties": {
"file": {
"type": "object",
"required": [
"name",
"resourceUri"
],
"properties": {
"name": {
"type": "string",
"maxLength": 512,
"minLength": 1,
"description": "The file name."
},
"resourceUri": {
"type": "string",
"maxLength": 2048,
"minLength": 1,
"description": "The file's resource URI from the host."
}
},
"description": "The file the host gave: its name and resourceUri.",
"additionalProperties": false
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"file": {
"type": "object",
"description": "The file's name and resourceUri."
},
"surface": {
"type": "string",
"description": "file."
}
},
"description": "The file the person opened, as a list."
},
"annotations": {
"title": "Open a list file",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description96/100
Searches the person's saved lists and records by name, for a mention in the composer. Returns up to 8 matches, each as a link with its name. Changes nothing.
The description explains the capability, when to use it, and what to expect.
Inputs94/100
Input documentation gives an agent enough context to populate the arguments.
query
Words in the list or record name. Empty gives recent ones.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Find a list or record to mention",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
Output schema72/100
Declare the result fields and their types in the output schema.
View output schema
{
"type": "object",
"properties": {
"items": {
"type": "array",
"items": {},
"description": "Each match as a resource link with its uri and name."
}
},
"description": "The saved lists and records whose names match the search."
}
Original tool metadata
{
"name": "search_lists",
"title": "Find a list or record to mention",
"description": "Searches the person's saved lists and records by name, for a mention in the composer. Returns up to 8 matches, each as a link with its name. Changes nothing.",
"inputSchema": {
"type": "object",
"required": [],
"properties": {
"query": {
"type": "string",
"default": "",
"maxLength": 200,
"description": "Words in the list or record name. Empty gives recent ones."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"items": {
"type": "array",
"items": {},
"description": "Each match as a resource link with its uri and name."
}
},
"description": "The saved lists and records whose names match the search."
},
"annotations": {
"title": "Find a list or record to mention",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
{
"name": "read_specstream_record_notes",
"title": "Read notes on a record",
"description": "Reads the notes on one record of a list, with the notes' version for the next change. Changes nothing.",
"inputSchema": {
"type": "object",
"title": "WorkspaceNoteScope",
"properties": {
"record_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Record Id",
"default": null,
"description": "The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"project_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id."
},
"collection_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Collection Id",
"default": null,
"description": "The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"notes": {
"type": "array",
"items": {}
},
"version": {
"type": "integer"
}
}
},
"annotations": {
"title": "Read notes on a record",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description85/100
Adds a note to one record of a list. An @ and a person's name mention that person. Returns the record's notes and their new version. The change fails if someone changed the notes after the latest read.
The description explains the capability, when to use it, and what to expect.
Inputs89/100
Remove repeated schema validation from descriptions; focus on meaning and usage context.
project_id
The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id.
Clear enough to supply this argument.
collection_id
The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id.
Clear enough to supply this argument.
record_id
The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id.
Clear enough to supply this argument.
expected_version
The version of the notes from the latest read of them, 0 when there are none. The change fails if someone changed the notes after that read.
Clear enough to supply this argument.
id
An id the caller chooses for the new note, unique among the notes on it.
Clear enough to supply this argument.
text
The note text. An @ and a person's name mention that person.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Add a note to a record",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema36/100
Describe the returned fields so an agent can interpret the result.
{
"name": "add_specstream_record_note",
"title": "Add a note to a record",
"description": "Adds a note to one record of a list. An @ and a person's name mention that person. Returns the record's notes and their new version. The change fails if someone changed the notes after the latest read.",
"inputSchema": {
"type": "object",
"title": "WorkspaceNoteAdd",
"required": [
"expected_version",
"id",
"text"
],
"properties": {
"id": {
"type": "string",
"title": "Id",
"maxLength": 128,
"minLength": 1,
"description": "An id the caller chooses for the new note, unique among the notes on it."
},
"text": {
"type": "string",
"title": "Text",
"maxLength": 8000,
"minLength": 1,
"description": "The note text. An @ and a person's name mention that person."
},
"record_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Record Id",
"default": null,
"description": "The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"project_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id."
},
"collection_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Collection Id",
"default": null,
"description": "The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"expected_version": {
"type": "integer",
"title": "Expected Version",
"minimum": 0,
"description": "The version of the notes from the latest read of them, 0 when there are none. The change fails if someone changed the notes after that read."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"notes": {
"type": "array",
"items": {}
},
"version": {
"type": "integer"
}
}
},
"annotations": {
"title": "Add a note to a record",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
Align the title with the capability actually described by this tool.
Description89/100
Marks a note on a record done, or opens it again. The note text stays. Returns the record's notes and their new version. The change fails if someone changed the notes after the latest read.
The description explains the capability, when to use it, and what to expect.
Inputs89/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
project_id
The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id.
Clear enough to supply this argument.
collection_id
The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
record_id
The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id.
Clear enough to supply this argument.
expected_version
The version of the notes from the latest read of them. The change fails if someone changed the notes after that read.
Clear enough to supply this argument.
resolved
true marks the note done. false opens it again.
Clear enough to supply this argument.
note_id
The note's id, from the notes on it.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Mark a record note done",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
Output schema34/100
Describe the returned fields so an agent can interpret the result.
{
"name": "resolve_specstream_record_note",
"title": "Mark a record note done",
"description": "Marks a note on a record done, or opens it again. The note text stays. Returns the record's notes and their new version. The change fails if someone changed the notes after the latest read.",
"inputSchema": {
"type": "object",
"title": "NoteResolve",
"required": [
"expected_version",
"resolved",
"note_id"
],
"properties": {
"note_id": {
"type": "string",
"title": "Note Id",
"maxLength": 128,
"minLength": 1,
"description": "The note's id, from the notes on it."
},
"resolved": {
"type": "boolean",
"title": "Resolved",
"description": "true marks the note done. false opens it again."
},
"record_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Record Id",
"default": null,
"description": "The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"project_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id."
},
"collection_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Collection Id",
"default": null,
"description": "The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"expected_version": {
"type": "integer",
"title": "Expected Version",
"minimum": 1,
"description": "The version of the notes from the latest read of them. The change fails if someone changed the notes after that read."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"notes": {
"type": "array",
"items": {}
},
"version": {
"type": "integer"
}
}
},
"annotations": {
"title": "Mark a record note done",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description90/100
Replaces the text of a note on a record. Its done state stays. Returns the record's notes and their new version. The change fails if someone changed the notes after the latest read.
The description explains the capability, when to use it, and what to expect.
Inputs89/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
project_id
The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id.
Clear enough to supply this argument.
collection_id
The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
record_id
The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
note_id
The note's id, from the notes on it.
Clear enough to supply this argument.
expected_version
The version of the notes from the latest read of them. The change fails if someone changed the notes after that read.
Clear enough to supply this argument.
text
The note text. An @ and a person's name mention that person.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Edit a record note",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
Output schema33/100
Describe the returned fields so an agent can interpret the result.
{
"name": "edit_specstream_record_note",
"title": "Edit a record note",
"description": "Replaces the text of a note on a record. Its done state stays. Returns the record's notes and their new version. The change fails if someone changed the notes after the latest read.",
"inputSchema": {
"type": "object",
"title": "NoteEdit",
"required": [
"note_id",
"expected_version",
"text"
],
"properties": {
"text": {
"type": "string",
"title": "Text",
"maxLength": 8000,
"minLength": 1,
"description": "The note text. An @ and a person's name mention that person."
},
"note_id": {
"type": "string",
"title": "Note Id",
"maxLength": 128,
"minLength": 1,
"description": "The note's id, from the notes on it."
},
"record_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Record Id",
"default": null,
"description": "The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"project_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id."
},
"collection_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Collection Id",
"default": null,
"description": "The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"expected_version": {
"type": "integer",
"title": "Expected Version",
"minimum": 1,
"description": "The version of the notes from the latest read of them. The change fails if someone changed the notes after that read."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"notes": {
"type": "array",
"items": {}
},
"version": {
"type": "integer"
}
}
},
"annotations": {
"title": "Edit a record note",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
}
Expand internal abbreviations and remove technical prefixes from the display title.
Description84/100
Deletes a note from a record. The other notes stay. Returns the record's notes and their new version. The change fails if someone changed the notes after the latest read.
The description explains the capability, when to use it, and what to expect.
Inputs83/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
project_id
The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id.
Clear enough to supply this argument.
collection_id
The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
record_id
The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
note_id
The note's id, from the notes on it.
Clear enough to supply this argument.
expected_version
The version of the notes from the latest read of them. The change fails if someone changed the notes after that read.
Clear enough to supply this argument.
Annotations100/100
All four behavior hints are declared.
{
"title": "Delete a record note",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
Output schema37/100
Describe the returned fields so an agent can interpret the result.
{
"name": "delete_specstream_record_note",
"title": "Delete a record note",
"description": "Deletes a note from a record. The other notes stay. Returns the record's notes and their new version. The change fails if someone changed the notes after the latest read.",
"inputSchema": {
"type": "object",
"title": "NoteDelete",
"required": [
"note_id",
"expected_version"
],
"properties": {
"note_id": {
"type": "string",
"title": "Note Id",
"maxLength": 128,
"minLength": 1,
"description": "The note's id, from the notes on it."
},
"record_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Record Id",
"default": null,
"description": "The record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"project_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Project Id",
"default": null,
"description": "The project, for a project note. A project note takes project_id. A record note takes collection_id and record_id."
},
"collection_id": {
"anyOf": [
{
"type": "string",
"maxLength": 128,
"minLength": 1
},
{
"type": "null"
}
],
"title": "Collection Id",
"default": null,
"description": "The list that holds the record, for a record note. A project note takes project_id. A record note takes collection_id and record_id."
},
"expected_version": {
"type": "integer",
"title": "Expected Version",
"minimum": 1,
"description": "The version of the notes from the latest read of them. The change fails if someone changed the notes after that read."
}
},
"additionalProperties": false
},
"outputSchema": {
"type": "object",
"properties": {
"notes": {
"type": "array",
"items": {}
},
"version": {
"type": "integer"
}
}
},
"annotations": {
"title": "Delete a record note",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
}
Loading score analytics…
Put your MCP to the test
A few tasks worth trying with this server, chosen from its reviewed tools. These have not been run.
01Search your workFits when the user asks for something already in their SpecStream account, by name or text.
02Find a list or record to mentionSearches the person's saved lists and records by name, for a mention in the composer.
03List projectsLists the projects the caller can open, one page at a time, with the caller's role in each.
04List collectionsFits when the user asks which lists they have.
05List saved requestsLists the saved requests the caller can open, one page at a time, with the id and current version of each.
Want to run these tasks against your MCP server with us?
SpecStream setup, tools, quality score, publishing readiness, and alternatives.
How do I add SpecStream to Claude, ChatGPT, or Cursor?
Add SpecStream to Claude, ChatGPT, or Cursor by connecting its remote MCP URL in each client's MCP settings. In Claude, add a custom connector. In ChatGPT, add an MCP app. In Cursor, add an MCP server under Settings. The MCP URL is https://api.specstream.ai/mcp.
Is SpecStream MCP ready for ChatGPT and Claude?
As of November 15, 2026, SpecStream scores 100/100 on Manufact's MCP publishing checklist. The scan meets ChatGPT publishing checks and meets Claude publishing checks. Passing these checks does not guarantee marketplace approval.
How clear is SpecStream's MCP tool documentation?
As of November 15, 2026, SpecStream scores 84/100 in Manufact's review of tool titles, descriptions, input and output schemas, and behavior annotations. The review covers 71 of 71 discovered tools. The quality score measures documentation clarity; runtime behavior and marketplace readiness require separate checks.
What are the best alternatives to SpecStream?
As of November 16, 2026, alternatives to SpecStream in search-research on the Manufact MCP leaderboard include TVFoodMaps, Search1API, US Internet Serviceability, Exa, and Italian Strikes.
How can SpecStream improve its MCP documentation?
get_reconciled_input is the lowest-scoring reviewed tool at 72/100. Describe the returned fields so an agent can interpret the result. Open its review for feedback and a copyable fix prompt, then rerun the documentation review after updating your tool metadata.
Not your MCP?
Review any public or OAuth MCP server for protocol compatibility, tool documentation, and ChatGPT and Claude publishing readiness. No account required.