Prov init v1. If the final line is not `ECO-WF-v1-END`, read
`ecco://workflow/full/v1`.
Prov records authority; agents execute. A claim never grants permission to
execute; nor does connection or assignment. Supervised work needs the person's
instruction and still obeys API denial, claims and policy.
First call `get_connection_context` for identity, workspace and current access.
Read-only cannot write. For a task call
`get_issue(identifier=..., view="compact")`; read parent and gates, then
`claim(start_work=false)`.
Before writes, recheck access and revision. Text edits need `base_revision`;
review/Done need evidence. Inspect errors and receipts; transport success is not
commit proof. Bulk rules: `ecco://workflow/bulk-write/v1`. Read before batch
writes. Board access grants no merge, deploy, spend or delete permission.
ECO-WF-v1-END
Discoverability
1/6 passed1 of 6 passed17/100
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 manifest declares this MCP server's exact endpoint.
Align the title with the capability actually described by this tool.
Description75/100
Acknowledge the current lease interruption after reviewing its feed.
`incorporated` and `superseded` resume governed work. The
`needs_clarification` and `cannot_incorporate` dispositions remain gated
and expose the note to the operator. Pass the `runner_id` you claimed
with (default: the connection's runner).
Explain mutations, irreversible effects, or prerequisites that matter before calling this tool.
Inputs0/100
0 of 6 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
ticket_ref
Not provided
Missing description. Explain what this value means and how to supply it.
acknowledgement_token
Not provided
Missing description. Explain what this value means and how to supply it.
disposition
Not provided
Missing description. Explain what this value means and how to supply it.
note
Not provided
Missing description. Explain what this value means and how to supply it.
runner_id
Not provided
Missing description. Explain what this value means and how to supply it.
executor
Not provided
Missing description. Explain what this value means and how to supply it.
Align the title with the capability actually described by this tool.
Description86/100
Append to or replace one canonical section of an issue body.
`issue` accepts a UUID or identifier such as ECO-947. Accepted sections
are Outcome, Scope, Decisions, Dependencies, and Proof (Definition of
Done); `Proof` is accepted as shorthand. The server changes only that
section and records the edit through normal issue revision history.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 6 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue
Not provided
Missing description. Explain what this value means and how to supply it.
section
Not provided
Missing description. Explain what this value means and how to supply it.
operation
Not provided
Missing description. Explain what this value means and how to supply it.
content
Not provided
Missing description. Explain what this value means and how to supply it.
base_revision
Not provided
Missing description. Explain what this value means and how to supply it.
sprint_id
Not provided
Missing description. Explain what this value means and how to supply it.
Shorten the title to an action and subject; move usage details into the description.
Description85/100
Archive an issue by id (UUID) or identifier (e.g. ECO-1).
Returns {identifier, archived: true} on success.
Returns structured error if the issue is not found.
Explain mutations, irreversible effects, or prerequisites that matter before calling this tool.
Inputs0/100
0 of 3 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
id
Not provided
Missing description. Explain what this value means and how to supply it.
identifier
Not provided
Missing description. Explain what this value means and how to supply it.
sprint_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
Attach a bounded typed artifact, with authenticated provenance.
Records submitted evidence only; never changes task state or grants
permission. Use design-doc, benchmark, test-output, tape, trace, or
decision-record. Existing reconciliation tools remain unchanged.
`artifact` is `{"kind": "test-output", "ref_or_blob": "<URL or up to
16 KB of text>", "provenance": {"observed_at": "<ISO time with zone>"}}`.
`criterion_refs` (ECO-1146) names the Proof items it covers (the
packet's `criterion_evidence[].source_ref`). Prov then records a
delivery receipt pointing at the artifact, and the next packet shows
those items as covered (submitted, not accepted).
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
artifact
Not provided
Missing description. Explain what this value means and how to supply it.
idempotency_key
Not provided
Missing description. Explain what this value means and how to supply it.
criterion_refs
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description86/100
Record a commit, PR, branch or session reference; never changes task state.
Pass a typed reference with kind and URL/hash/ref, plus optional metadata.
`link.kind` is one of "pr", "commit", "branch" or "session", for example
`{"kind": "pr", "url": "https://github.com/org/repo/pull/42"}`.
Authorship is derived from the authenticated connection, never tool input.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 6 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › TaskLinkInput › kind
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › TaskLinkInput › url
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › TaskLinkInput › hash
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › TaskLinkInput › ref
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › TaskLinkInput › metadata
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Atomically archive or unarchive 1..2000 revisioned tasks.
`manifest_digest` is SHA-256 over canonical JSON with sorted object keys
and compact separators: `{"operation":...,"items":[{"identifier":...,
"base_revision":...}]}` in the submitted order, prefixed by `sha256:`.
The entire manifest is authorized, locked, revision-checked, and
validated before any task changes. Any missing, hidden, duplicated,
stale, or already-in-target-state task aborts the operation. Exact
idempotent retries return the stored result.
The default receipt is aggregate-only and binds the exact set by its
digest. Set `include_items=true` when per-task protocol results are
needed and provide a sufficient response budget. Existing task change
history retains one revision and event per changed task either way.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 7 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
operation
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchLifecycleItem › identifier
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchLifecycleItem › base_revision
Not provided
Missing description. Explain what this value means and how to supply it.
manifest_digest
Not provided
Missing description. Explain what this value means and how to supply it.
idempotency_key
Not provided
Missing description. Explain what this value means and how to supply it.
include_items
Not provided
Missing description. Explain what this value means and how to supply it.
max_bytes
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
Atomically update 1..25 selected tasks in one deterministic call.
Each input carries one task identifier, its positive `base_revision`,
and all known changes for that task. Results preserve input order. The
batch commits only when every task passes the existing single-task
authorization, validation, claim and transition rules; otherwise every
item reports conflict/error/aborted and no task changes. An exact retry
with the same idempotency key returns the stored receipt. Reusing a key
with different input is refused. The response budget covers both MCP
channels and any passive task-change notice; retry a budget rejection
with a larger value because no mutation was committed.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 27 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
$defs › BatchUpdateItem › identifier
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateItem › base_revision
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › title
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › description
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › state
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › priority
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › estimate
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › assignee_id
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › status_id
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › labels
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › due_date
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › branch
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › suite_delta
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › actuals
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › commit
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › review_state
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › pr_url
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › gates
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › checklist_item
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › depends_on
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › executor
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › outcome
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › metadata
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › kind
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › BatchUpdateChanges › status_note
Not provided
Missing description. Explain what this value means and how to supply it.
idempotency_key
Not provided
Missing description. Explain what this value means and how to supply it.
max_bytes
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description90/100
Save 1–100 authored lines to one existing project as To do tasks.
Use a project UUID or exact name and retain request_key across retries.
Preview writes nothing. No decomposition, project creation or execution
approval occurs. Returns the original committed receipt on exact retry.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
text
Not provided
Missing description. Explain what this value means and how to supply it.
project
Not provided
Missing description. Explain what this value means and how to supply it.
request_key
Not provided
Missing description. Explain what this value means and how to supply it.
preview
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description98/100
Claim ownership; start_work=false preserves progress.
Legacy callers default to claim-and-start (InProgress). New coordinators
should pass start_work=false and record an authorized start separately.
Retrying a live claim never starts work as a side effect.
The owner is this connection's actor plus ``runner_id``. Subagents that
share one connection each pass their own ``runner_id`` and compete for
the claim exactly as separate connections do: a different runner_id
while a live lease exists gets ``claimed: false`` naming the live owner
(``current_actor_id``, ``current_runner_id``, ``current_executor``); the
same runner_id and executor retry is ``idempotent: true``. ``executor``
and ``runner_id`` default to the connection's values when omitted, and
the response echoes what you sent. A malformed value is refused
(``invalid_runner_id``/``invalid_executor``: 1-128/1-64 chars of
``A-Za-z0-9._:@/+=-``), never rewritten. runner_id is not an
authorization source; actor and access stay server-derived.
``owner_ref`` + ``idempotency_key`` is accepted for durable supervisor
callers (owner_ref defaults to runner_id). A replay with the same pair
returns the existing lease only for the same actor and runner_id; the
key never moves a live claim between runners.
Ownership alone grants no execution permission. MCP workers cannot issue
execution approval, including through the ``authorization`` argument.
Ask a human to approve the current task context when required, then
re-read execution_packet.ready. Task-management access is not approval.
Use view="compact" for one copy of the task prose, the post-claim next
action and current gates. The default full packet remains available to
existing callers; compact includes links to its source reads.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 9 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id_or_identifier
Not provided
Missing description. Explain what this value means and how to supply it.
executor
Not provided
Missing description. Explain what this value means and how to supply it.
runner_id
Not provided
Missing description. Explain what this value means and how to supply it.
owner_ref
Not provided
Missing description. Explain what this value means and how to supply it.
idempotency_key
Not provided
Missing description. Explain what this value means and how to supply it.
ttl_sec
Not provided
Missing description. Explain what this value means and how to supply it.
authorization
Not provided
Missing description. Explain what this value means and how to supply it.
start_work
Not provided
Missing description. Explain what this value means and how to supply it.
view
Not provided
Missing description. Explain what this value means and how to supply it.
{
"name": "claim",
"title": "Claim task ownership",
"description": "Claim ownership; start_work=false preserves progress.\n\n Legacy callers default to claim-and-start (InProgress). New coordinators\n should pass start_work=false and record an authorized start separately.\n Retrying a live claim never starts work as a side effect.\n\n The owner is this connection's actor plus ``runner_id``. Subagents that\n share one connection each pass their own ``runner_id`` and compete for\n the claim exactly as separate connections do: a different runner_id\n while a live lease exists gets ``claimed: false`` naming the live owner\n (``current_actor_id``, ``current_runner_id``, ``current_executor``); the\n same runner_id and executor retry is ``idempotent: true``. ``executor``\n and ``runner_id`` default to the connection's values when omitted, and\n the response echoes what you sent. A malformed value is refused\n (``invalid_runner_id``/``invalid_executor``: 1-128/1-64 chars of\n ``A-Za-z0-9._:@/+=-``), never rewritten. runner_id is not an\n authorization source; actor and access stay server-derived.\n\n ``owner_ref`` + ``idempotency_key`` is accepted for durable supervisor\n callers (owner_ref defaults to runner_id). A replay with the same pair\n returns the existing lease only for the same actor and runner_id; the\n key never moves a live claim between runners.\n Ownership alone grants no execution permission. MCP workers cannot issue\n execution approval, including through the ``authorization`` argument.\n Ask a human to approve the current task context when required, then\n re-read execution_packet.ready. Task-management access is not approval.\n Use view=\"compact\" for one copy of the task prose, the post-claim next\n action and current gates. The default full packet remains available to\n existing callers; compact includes links to its source reads.\n ",
"inputSchema": {
"type": "object",
"title": "claimArguments",
"required": [
"issue_id_or_identifier"
],
"properties": {
"view": {
"enum": [
"full",
"compact"
],
"type": "string",
"title": "View",
"default": "full"
},
"ttl_sec": {
"type": "integer",
"title": "Ttl Sec",
"default": 300
},
"executor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Executor",
"default": null
},
"owner_ref": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Owner Ref",
"default": null
},
"runner_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Runner Id",
"default": null
},
"start_work": {
"type": "boolean",
"title": "Start Work",
"default": true
},
"authorization": {
"anyOf": [
{
"type": "object",
"additionalProperties": true
},
{
"type": "null"
}
],
"title": "Authorization",
"default": null
},
"idempotency_key": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Idempotency Key",
"default": null
},
"issue_id_or_identifier": {
"type": "string",
"title": "Issue Id Or Identifier"
}
}
},
"outputSchema": {
"type": "object",
"title": "claimDictOutput",
"additionalProperties": true
},
"annotations": {
"title": "Claim task ownership",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description81/100
Create a comment; kind=decision plus options=[] raises a request;
kind=review plus review={"type": ...} records a typed review fact
(ECO-818: review-requested, ownership-assigned/accepted/rejected,
finding, fix-submitted, re-review).
Writing a decision (ECO-1181/1182): put the ask first, in one or two
sentences a person can answer from the Inbox card, and keep the body
under about 600 characters. Put proof, test counts and deploy notes in
a separate note comment or the PR, and link them. Give options as
`{"label": ..., "why": ...}`: for your pick, why you recommend it; for
each other option, why not (one line each). Name your pick in
`recommendation` by option number (1-based) or exact label; it's
preselected in the Inbox, and a pick that matches no option is
rejected.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 12 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
body
Not provided
Missing description. Explain what this value means and how to supply it.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
author
Not provided
Missing description. Explain what this value means and how to supply it.
issue_identifier
Not provided
Missing description. Explain what this value means and how to supply it.
kind
Not provided
Missing description. Explain what this value means and how to supply it.
payload
Not provided
Missing description. Explain what this value means and how to supply it.
options
Not provided
Missing description. Explain what this value means and how to supply it.
blocks
Not provided
Missing description. Explain what this value means and how to supply it.
blocking_condition
Not provided
Missing description. Explain what this value means and how to supply it.
reason
Not provided
Missing description. Explain what this value means and how to supply it.
recommendation
Not provided
Missing description. Explain what this value means and how to supply it.
review
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
Create a task in a project identified by UUID or exact name.
Write an OUTCOME-FIRST title; put implementation details in body.
validate_only=true returns {"valid": true} or a validation error without
creating a task or consuming an identifier. parent_id accepts UUID or
identifier and must reference a top-level task (depth-1 only).
labels accepts existing label names (case-insensitive, in the task's
team) or label UUIDs; label_ids accepts UUIDs. labels never creates a
label: an unknown entry refuses the whole create with error
"unknown_label", did_you_mean close names and the create_label call to
make first. Labels commit with creation or roll back together. Exact
keyed retries return the original task and labels; changed labels with
the same key conflict. Receipt metadata preserves the original
authenticated actor and creation facts.
sprint_id records intent only; reserve_for_sprint is a separate call.
due_date is ISO-8601. Optional acceptance, constraints, file_pointers,
branch and suggested_model store the authored build specification.
view="compact" returns only identity, state and revision instead of
echoing the created task; the default "full" is unchanged.
Follow the workspace defaults from get_connection_context; a task that
misses one is still created and returns ``advice`` to act on or relay to
your person. Only an immediate retry of the same title in the project is
refused, unless confirm_new=true or idempotency_key. kind is optional:
feature, integration, refactor, domain, security, docs or chore.
metadata may be partial (e.g. {"evidence": [...]}); disposition and
next_action are derived from state.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 24 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
title
Not provided
Missing description. Explain what this value means and how to supply it.
project_id
Not provided
Missing description. Explain what this value means and how to supply it.
body
Not provided
Missing description. Explain what this value means and how to supply it.
state
Not provided
Missing description. Explain what this value means and how to supply it.
priority
Not provided
Missing description. Explain what this value means and how to supply it.
estimate
Not provided
Missing description. Explain what this value means and how to supply it.
assignee_id
Not provided
Missing description. Explain what this value means and how to supply it.
parent_id
Not provided
Missing description. Explain what this value means and how to supply it.
status_id
Not provided
Missing description. Explain what this value means and how to supply it.
label_ids
Not provided
Missing description. Explain what this value means and how to supply it.
labels
Not provided
Missing description. Explain what this value means and how to supply it.
sprint_id
Not provided
Missing description. Explain what this value means and how to supply it.
due_date
Not provided
Missing description. Explain what this value means and how to supply it.
branch
Not provided
Missing description. Explain what this value means and how to supply it.
validate_only
Not provided
Missing description. Explain what this value means and how to supply it.
idempotency_key
Not provided
Missing description. Explain what this value means and how to supply it.
depends_on
Not provided
Missing description. Explain what this value means and how to supply it.
executor
Not provided
Missing description. Explain what this value means and how to supply it.
outcome
Not provided
Missing description. Explain what this value means and how to supply it.
metadata
Not provided
Missing description. Explain what this value means and how to supply it.
metadata_hints
Not provided
Missing description. Explain what this value means and how to supply it.
view
Not provided
Missing description. Explain what this value means and how to supply it.
kind
Not provided
Missing description. Explain what this value means and how to supply it.
confirm_new
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
Create a new standalone label. Returns the created label object.
Returns 409-like error if a label with the same name already exists
within the team. If team_id is omitted, defaults to the first available team.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 3 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
name
Not provided
Missing description. Explain what this value means and how to supply it.
color
Not provided
Missing description. Explain what this value means and how to supply it.
team_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description89/100
Create a new project under an existing or a new initiative.
Pass exactly one destination: `initiative_id` for an existing
initiative, or `new_initiative` ({name, description?, team_id?}) to
create the initiative and the project together in one transaction.
Without `new_initiative.team_id` the initiative goes to this agent's
home team, as with `create_initiative`; a retry with the same
`idempotency_key` after that home team changes returns
`idempotency_conflict` instead of creating a second pair.
Errors: `not_found` if the initiative or team is not visible to this
connection; `team_limited` if this connection is limited to its home
team and the destination is another team; `invalid_request` for a
missing, doubled or invalid destination; `permission_required` if
this connection cannot create projects.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 8 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
name
Not provided
Missing description. Explain what this value means and how to supply it.
initiative_id
Not provided
Missing description. Explain what this value means and how to supply it.
description
Not provided
Missing description. Explain what this value means and how to supply it.
status_id
Not provided
Missing description. Explain what this value means and how to supply it.
idempotency_key
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › ProjectNewInitiative › name
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › ProjectNewInitiative › description
Not provided
Missing description. Explain what this value means and how to supply it.
$defs › ProjectNewInitiative › team_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description94/100
Create a relation edge between two issues.
`type` must be 'blocks' or 'relates_to'. A 'blocks' edge means
source_issue_id blocks target_issue_id (target is blocked_by source).
Returns the created relation on success, or an error envelope on
duplicate/self-edge/not-found.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 3 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
source_issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
target_issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
type
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description91/100
Create a new team.
``prefix`` must be uppercase ASCII letters only (e.g. ``ECO``, ``PLAT``).
Returns the created team object.
Returns ``{"error": "conflict", "detail": "..."}`` if a team with the
same prefix already exists.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 3 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
name
Not provided
Missing description. Explain what this value means and how to supply it.
prefix
Not provided
Missing description. Explain what this value means and how to supply it.
description
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Hard-delete a comment by UUID.
Returns `{"deleted": true}` on success.
Returns `{"deleted": false, "error": "not_found"}` if the comment does
not exist.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
comment_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description98/100
Hard-delete an initiative by UUID. Cascades to projects, issues.
Returns `{"deleted": true}` on success.
Returns `{"deleted": false, "error": "not_found"}` if the initiative
does not exist.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
initiative_id
Not provided
Missing description. Explain what this value means and how to supply it.
Align the title with the capability actually described by this tool.
Description89/100
Hard-delete an issue by UUID (`issue_id`) or identifier (`identifier`),
cascading to comments, labels, and reservations.
Returns `{"deleted": true}` on success.
Returns `{"deleted": false, "error": "not_found"}` if the issue does not exist.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 3 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
identifier
Not provided
Missing description. Explain what this value means and how to supply it.
sprint_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
Delete a label by ID.
Returns `{"deleted": true}` on success.
Returns `{"deleted": false, "error": "not_found"}` if the label
does not exist.
Returns `{"deleted": false, "error": "referenced"}` if any issue
still references this label.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description97/100
Hard-delete a project by UUID. Cascades to issues.
Returns `{"deleted": true}` on success.
Returns `{"deleted": false, "error": "not_found"}` if the project
does not exist.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
project_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Delete a status by ID.
Returns `{"deleted": true}` on success.
Returns `{"deleted": false, "error": "not_found"}` if the status
does not exist.
Returns `{"deleted": false, "error": "referenced"}` if any issue
still references this status.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description94/100
Delete a team by UUID unless retained references prevent deletion.
Returns `{"deleted": true}` on success.
Returns `{"deleted": false, "error": "not_found"}` if the team does
not exist.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
team_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description79/100
Ask an operator for a decision on an accessible task.
``proposal`` is required so the ask reaches Overview with a concrete
next action rather than making the operator reconstruct one. Worker
credentials may only target tasks inside their existing project grant.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 8 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
app_id
Not provided
Missing description. Explain what this value means and how to supply it.
issue_ref
Not provided
Missing description. Explain what this value means and how to supply it.
reason
Not provided
Missing description. Explain what this value means and how to supply it.
proposal
Not provided
Missing description. Explain what this value means and how to supply it.
severity
Not provided
Missing description. Explain what this value means and how to supply it.
warning_class
Not provided
Missing description. Explain what this value means and how to supply it.
evidence_refs
Not provided
Missing description. Explain what this value means and how to supply it.
snapshot_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description91/100
Record that an agent is blocked on this issue. Resolve by UUID
(`issue_id`) or identifier (`identifier`, e.g. ECO-1).
Requires an active blocking relation or applicable open typed decision.
Atomically sets Blocked and status_note, and posts a [blocked] comment.
Returns the updated issue + comment.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 5 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
identifier
Not provided
Missing description. Explain what this value means and how to supply it.
reason
Not provided
Missing description. Explain what this value means and how to supply it.
blocker_class
Not provided
Missing description. Explain what this value means and how to supply it.
sprint_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description84/100
Flag an issue as needing human review. Resolve by UUID (`issue_id`)
or identifier (`identifier`, e.g. ECO-1).
Records a comment with the reason and blocker_class, and returns
the updated issue + comment.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 5 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
identifier
Not provided
Missing description. Explain what this value means and how to supply it.
reason
Not provided
Missing description. Explain what this value means and how to supply it.
blocker_class
Not provided
Missing description. Explain what this value means and how to supply it.
sprint_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description97/100
Return the authenticated worker's identity, workspace and current access.
Includes onboarding information, this workspace's ``defaults`` for
using the board (follow them; ask your person when one doesn't fit)
and available capabilities. ``purpose``, when set, is what your person
said this agent is for: let it guide where you focus. Optional
model records a self-reported display label for this worker on HTTP
connections; it does not change permissions. With the human's direction,
request_access="manage_tasks" requests an OAuth consent flow for task
management. Access changes require human approval. Paired connections
require workspace-administrator approval instead.
The default detail="full" keeps the complete first-session response.
For repeat checks, pass detail="compact" and the last
known_defaults_version. Compact returns the effective workspace
defaults text only when its version changed; it skips welcome and
optional board provenance reads but retains access and recovery facts.
A truncated purpose or selected-project list is labeled as a preview
with counts and a full-detail retrieval path.
Arguments cannot select another worker or workspace. This tool does not
change tasks or authorize execution.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
request_access
Not provided
Missing description. Explain what this value means and how to supply it.
model
Not provided
Missing description. Explain what this value means and how to supply it.
detail
Not provided
Missing description. Explain what this value means and how to supply it.
known_defaults_version
Not provided
Missing description. Explain what this value means and how to supply it.
{
"name": "get_connection_context",
"title": "Get connection context",
"description": "Return the authenticated worker's identity, workspace and current access.\n\n Includes onboarding information, this workspace's ``defaults`` for\n using the board (follow them; ask your person when one doesn't fit)\n and available capabilities. ``purpose``, when set, is what your person\n said this agent is for: let it guide where you focus. Optional\n model records a self-reported display label for this worker on HTTP\n connections; it does not change permissions. With the human's direction,\n request_access=\"manage_tasks\" requests an OAuth consent flow for task\n management. Access changes require human approval. Paired connections\n require workspace-administrator approval instead.\n The default detail=\"full\" keeps the complete first-session response.\n For repeat checks, pass detail=\"compact\" and the last\n known_defaults_version. Compact returns the effective workspace\n defaults text only when its version changed; it skips welcome and\n optional board provenance reads but retains access and recovery facts.\n A truncated purpose or selected-project list is labeled as a preview\n with counts and a full-detail retrieval path.\n Arguments cannot select another worker or workspace. This tool does not\n change tasks or authorize execution.\n ",
"inputSchema": {
"type": "object",
"title": "get_connection_contextArguments",
"properties": {
"model": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Model",
"default": null
},
"detail": {
"enum": [
"full",
"compact"
],
"type": "string",
"title": "Detail",
"default": "full"
},
"request_access": {
"anyOf": [
{
"type": "string",
"const": "manage_tasks"
},
{
"type": "null"
}
],
"title": "Request Access",
"default": null
},
"known_defaults_version": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Known Defaults Version",
"default": null
}
}
},
"outputSchema": {
"type": "object",
"title": "get_connection_contextDictOutput",
"additionalProperties": true
},
"annotations": {
"title": "Get connection context",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
}
Align the title with the capability actually described by this tool.
Description93/100
With read_contract="compact.v1", omitted view selects compact.
Omitted/legacy contract selects full; explicit view always wins.
Unsupported contract versions are refused.
Fetch a single issue by UUID (`id`) or identifier (`identifier`, e.g. ECO-1),
including nested labels and comments.
activity_cursor=true includes a scoped activity fingerprint. Recheck with
since_cursor; unchanged covers task, comments and accepted-brief sources.
Incomplete sources never certify unchanged. Changed output is bounded
with explicit full-context retrieval. Existing since_revision is unchanged.
view="resume" returns one bounded read of the current brief, live claim,
gates/blockers, recorded tested commit and next proof. Unknown evidence
stays unknown; this read never claims or grants execution permission.
view="compact" returns a short state header, the description and the
latest `comments` (default 2, max 20, each capped at 600 chars) with the
list_comments offset for older ones. Newer comments, child activity and
recorded delivery references produce source-linked freshness cues, not
inferred task state; empty fields are omitted. It is
markdown by default, or the same fields as a dict with format="json".
With `since_revision`, an unchanged task returns only its revision,
state and comment count. The default "full" view is unchanged.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 9 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
id
Not provided
Missing description. Explain what this value means and how to supply it.
identifier
Not provided
Missing description. Explain what this value means and how to supply it.
view
Not provided
Missing description. Explain what this value means and how to supply it.
read_contract
Not provided
Missing description. Explain what this value means and how to supply it.
format
Not provided
Missing description. Explain what this value means and how to supply it.
comments
Not provided
Missing description. Explain what this value means and how to supply it.
since_revision
Not provided
Missing description. Explain what this value means and how to supply it.
activity_cursor
Not provided
Missing description. Explain what this value means and how to supply it.
since_cursor
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Read selected tasks in input order within one total response budget.
Pass 1..50 unique UUIDs or task identifiers. Every visible item always
carries identifier, state and revision; `fields` selects documented
compact-detail fields. Missing and inaccessible tasks share the same
not_found receipt. Whole items are paged, never clipped; pass the signed
`next_cursor` with the identical request to continue. An oversized item
is explicit and points back to get_issue. Authorization is rechecked for
every task and every page. For cheap rechecks, map task references to
their last seen positive revisions in `since_revisions`; unchanged rows
return comment freshness without resending full task content.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 6 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
identifiers
Not provided
Missing description. Explain what this value means and how to supply it.
fields
Not provided
Missing description. Explain what this value means and how to supply it.
max_bytes
Not provided
Missing description. Explain what this value means and how to supply it.
cursor
Not provided
Missing description. Explain what this value means and how to supply it.
comments
Not provided
Missing description. Explain what this value means and how to supply it.
since_revisions
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description96/100
Renew your live task claim with issue_id_or_identifier and ttl_sec.
The owner is this connection's actor plus runner_id. Pass the
``runner_id`` you claimed with (default: the connection's runner);
``executor`` is optional and, when given, must match the claim. A
sibling subagent on the same connection cannot renew another runner's
lease. Returns refreshed=false
with lease_not_held_or_expired when renewal is refused; re-read context
and coordinate, never take over another worker. Renewal does not extend
operator authorization. Without a task identifier, app_id and runner_id
select the legacy runner-liveness mode, which does not own a task and
stays bound to the connection.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 6 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
runner_id
Not provided
Missing description. Explain what this value means and how to supply it.
app_id
Not provided
Missing description. Explain what this value means and how to supply it.
ttl_sec
Not provided
Missing description. Explain what this value means and how to supply it.
issue_id_or_identifier
Not provided
Missing description. Explain what this value means and how to supply it.
owner_ref
Not provided
Missing description. Explain what this value means and how to supply it.
executor
Not provided
Missing description. Explain what this value means and how to supply it.
Align the title with the capability actually described by this tool.
Description97/100
Compose an at-a-glance summary of an issue from its related data.
Returns one object combining core fields (id, identifier, title,
state), comment_count (total), labels (list of label name strings),
watcher_count, reactions (a tally grouped by emoji, e.g.
{"👍": 3, "🎉": 1}), and latest_comment (the most recent comment's
body/author/created_at, or null when the issue has no comments).
An issue with zero comments/labels/watchers/reactions returns
zeros/empty collections and latest_comment=null.
Returns a not_found envelope when the issue does not exist.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
{
"name": "issue_digest",
"title": "Task digest",
"description": "Compose an at-a-glance summary of an issue from its related data.\n\n Returns one object combining core fields (id, identifier, title,\n state), comment_count (total), labels (list of label name strings),\n watcher_count, reactions (a tally grouped by emoji, e.g.\n {\"👍\": 3, \"🎉\": 1}), and latest_comment (the most recent comment's\n body/author/created_at, or null when the issue has no comments).\n\n An issue with zero comments/labels/watchers/reactions returns\n zeros/empty collections and latest_comment=null.\n Returns a not_found envelope when the issue does not exist.\n ",
"inputSchema": {
"type": "object",
"title": "issue_digestArguments",
"required": [
"issue_id"
],
"properties": {
"issue_id": {
"type": "string",
"title": "Issue Id"
}
}
},
"outputSchema": {
"type": "object",
"title": "issue_digestDictOutput",
"additionalProperties": true
},
"annotations": {
"title": "Task digest",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description98/100
List tasks completed (moved to Done) in a time window.
`since`/`until` are ISO-8601 timestamps with timezone; both optional —
the default window is the trailing 7 days ending now. Returns
{since, until, items, totals}: each item carries the Done-transition
timestamp and actor (change-record attribution), the task core fields,
reported executor, and ECO-765 actuals when the executor reported
them; totals roll up actuals across reporting tasks only. Read-only —
mirrors GET /reports/completed.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 2 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
since
Not provided
Missing description. Explain what this value means and how to supply it.
until
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description84/100
List bounded task and effective-epic changes since this agent's claim.
Omit `since_revision` to use the exact claim-time revision baselines. A
truncated response reports both the returned prefix and total count.
The claim is this connection's actor plus runner_id: pass the
`runner_id` you claimed with (default: the connection's runner).
Another runner's claim answers lease_not_held.
Describe the returned information so the agent knows whether it will answer the request.
Inputs0/100
0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
ticket_ref
Not provided
Missing description. Explain what this value means and how to supply it.
since_revision
Not provided
Missing description. Explain what this value means and how to supply it.
runner_id
Not provided
Missing description. Explain what this value means and how to supply it.
executor
Not provided
Missing description. Explain what this value means and how to supply it.
Align the title with the capability actually described by this tool.
Description85/100
Read issue-scoped evidence with concise summaries. Set include_details
for full structured check records, commands and provenance. It is read-only,
and cannot approve, spend, or close work. The handler preserves
workspace scope for authenticated agents. Stop and seek an operator if
evidence would require broader access. Implementation uses the
connection-level supervised/unattended workflow.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
open_tasks_with_strong_completion
Not provided
Missing description. Explain what this value means and how to supply it.
limit
Not provided
Missing description. Explain what this value means and how to supply it.
include_details
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description91/100
With read_contract="compact.v1", omitted view selects compact.
Omitted/legacy contract selects full; explicit view always wins.
Unsupported contract versions are refused.
List issues. `state` accepts open,archived,deleted,terminal,any.
`open` = queueable (Backlog/Todo/InProgress, not deleted/archived).
`terminal` = finished (Done/Cancelled).
`label` filters by attached label.
`priority` filters by one or more priorities (1,2,3,4).
`order` controls sort: created_at_asc, created_at_desc, updated_at_asc,
updated_at_desc, priority_asc, priority_desc.
`soft_deleted` / `archived` control soft-delete/archive visibility
(false, true, or any).
`exclude_blocked` excludes issues blocked by non-terminal issues.
`ready=true` additionally excludes terminal, archived and deleted tasks.
Dependency readiness is not execution authorization.
`pickup=true` selects dependency-unblocked open work and decisions but
omits automatic Getting started milestones. Parent coordination remains
eligible; inspect `pickup` and `next_action` on each row. Pickup does not
grant execution authorization.
`only_blocked` returns only issues that are blocked by non-terminal issues.
`needs_human` filters by needs_human flag:
true=only flagged, false=only unflagged.
`has_label` filters by label presence (true=labelled, false=unlabelled).
`project_id` filters by project UUID.
`workflow_state` filters by exact workflow state column value; repeatable.
Allowed: Backlog, NeedsGrooming, Todo, InProgress, Blocked, AwaitingReview,
Done, Cancelled. `progress` filters To do (Todo), InProgress, Done or
Cancelled using recorded history, including legacy waiting tasks.
DISTINCT from `state` which controls visibility (open/archived/deleted/any).
`assignee_id` filters by assignee UUID.
`created_before` / `created_after` filter by creation timestamp (ISO-8601,
e.g. 2025-01-01T00:00:00).
`due_before` / `due_after` filter by due_date (ISO-8601); issues without a
due_date are excluded when either bound is set.
`parent_id` filters by parent issue UUID (returns child sub-issues).
`top_level_only` returns only root-level issues (no parent).
`view` selects the response shape and defaults to `full` (unchanged).
`view="compact"` returns only identity, state, priority, needs_human,
revision, updated_at, project and next_action per task — roughly a
twentieth of the bytes — plus a `reason_catalog` mapping each
next_action code to one sentence, held once per response instead of
once per row. Use it to pick the next task, then read that task with
`get_issue(view="compact")` for full detail; scope enforcement and
ordering are identical either way. A compact page is bounded by
`max_bytes`: when the budget is reached the page returns fewer whole
rows and reports `truncated` with a `next_offset` to resume from. No
row is ever partially returned and no field is truncated.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 27 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
state
Not provided
Missing description. Explain what this value means and how to supply it.
label
Not provided
Missing description. Explain what this value means and how to supply it.
priority
Not provided
Missing description. Explain what this value means and how to supply it.
order
Not provided
Missing description. Explain what this value means and how to supply it.
soft_deleted
Not provided
Missing description. Explain what this value means and how to supply it.
archived
Not provided
Missing description. Explain what this value means and how to supply it.
exclude_blocked
Not provided
Missing description. Explain what this value means and how to supply it.
only_blocked
Not provided
Missing description. Explain what this value means and how to supply it.
needs_human
Not provided
Missing description. Explain what this value means and how to supply it.
has_label
Not provided
Missing description. Explain what this value means and how to supply it.
project_id
Not provided
Missing description. Explain what this value means and how to supply it.
workflow_state
Not provided
Missing description. Explain what this value means and how to supply it.
progress
Not provided
Missing description. Explain what this value means and how to supply it.
assignee_id
Not provided
Missing description. Explain what this value means and how to supply it.
created_before
Not provided
Missing description. Explain what this value means and how to supply it.
created_after
Not provided
Missing description. Explain what this value means and how to supply it.
due_before
Not provided
Missing description. Explain what this value means and how to supply it.
due_after
Not provided
Missing description. Explain what this value means and how to supply it.
parent_id
Not provided
Missing description. Explain what this value means and how to supply it.
top_level_only
Not provided
Missing description. Explain what this value means and how to supply it.
limit
Not provided
Missing description. Explain what this value means and how to supply it.
offset
Not provided
Missing description. Explain what this value means and how to supply it.
ready
Not provided
Missing description. Explain what this value means and how to supply it.
pickup
Not provided
Missing description. Explain what this value means and how to supply it.
view
Not provided
Missing description. Explain what this value means and how to supply it.
read_contract
Not provided
Missing description. Explain what this value means and how to supply it.
max_bytes
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description91/100
List labels with optional team filtering and offset-based pagination.
When team_id is provided, returns labels belonging to that team
or global labels (team_id IS NULL).
Returns {"items": [LabelDetail, ...], "total_count": int}.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 3 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
team_id
Not provided
Missing description. Explain what this value means and how to supply it.
limit
Not provided
Missing description. Explain what this value means and how to supply it.
offset
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description94/100
List all relation edges for an issue (both directions).
Returns `{"items": [...], "total_count": N}` with every `blocks` and
`relates_to` edge where the issue appears as source or target.
Returns a not_found error envelope if the issue does not exist.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
List watchers for an issue.
Returns the watcher rows surfaced by `ecco.issues.repo.list_watchers`
— `{items: [{user_id, name, email, created_at}, ...], total_count}`.
Returns a `{"error": "not_found", "detail": ...}` envelope when the
issue does not exist (same pattern as list_comments).
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 1 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description88/100
Append minimal issue evidence through the existing ledger. This
does not authorize spending or change task state: external agents may
submit assigned audit-packet evidence or task_history delivery receipts
with the eco.delivery.v1 extension, without starting an audit run.
Operators submit operator notes. Existing task scope, synchronization
and secret boundaries apply; a receipt is reported evidence, not an
execution grant or independent verification.
`criterion_refs` (ECO-1146) names the Proof items this evidence covers:
the `criterion_evidence[].source_ref` values from the execution packet
(`issue:<id>:criterion:<digest>`). They're added to the eco.delivery.v1
receipt, so the next packet shows those items as covered (submitted,
not accepted). A delivery receipt is required to carry them.
Describe the returned information so the agent knows whether it will answer the request.
Inputs0/100
0 of 3 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
evidence
Not provided
Missing description. Explain what this value means and how to supply it.
criterion_refs
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description98/100
Release a lease idempotently, only if this actor's runner holds it.
Pass the ``lease_id``, ``app_id`` and ``runner_id`` from your claim or
heartbeat response. The owner is this connection's actor plus
runner_id (and app_id, the claim's executor): a sibling subagent on the
same connection cannot release another runner's lease, and neither can
another connection. Returns the released lease, or not_found when the
lease is unknown or not held by that runner. With expiry_only, release
only when the datastore judges it expired; a live lease returns an
explicit no-op without mutation evidence. Malformed values are refused
with invalid_app_id/invalid_runner_id, never rewritten.
Resolve contradictions between the description, input schema, and annotations.
Inputs0/100
0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
lease_id
Not provided
Missing description. Explain what this value means and how to supply it.
app_id
Not provided
Missing description. Explain what this value means and how to supply it.
runner_id
Not provided
Missing description. Explain what this value means and how to supply it.
expiry_only
Not provided
Missing description. Explain what this value means and how to supply it.
{
"name": "release_lease",
"title": "Release lease",
"description": "Release a lease idempotently, only if this actor's runner holds it.\n\n Pass the ``lease_id``, ``app_id`` and ``runner_id`` from your claim or\n heartbeat response. The owner is this connection's actor plus\n runner_id (and app_id, the claim's executor): a sibling subagent on the\n same connection cannot release another runner's lease, and neither can\n another connection. Returns the released lease, or not_found when the\n lease is unknown or not held by that runner. With expiry_only, release\n only when the datastore judges it expired; a live lease returns an\n explicit no-op without mutation evidence. Malformed values are refused\n with invalid_app_id/invalid_runner_id, never rewritten.\n ",
"inputSchema": {
"type": "object",
"title": "release_leaseArguments",
"required": [
"lease_id",
"app_id",
"runner_id"
],
"properties": {
"app_id": {
"type": "string",
"title": "App Id"
},
"lease_id": {
"type": "string",
"title": "Lease Id"
},
"runner_id": {
"type": "string",
"title": "Runner Id"
},
"expiry_only": {
"type": "boolean",
"title": "Expiry Only",
"default": false
}
}
},
"outputSchema": {
"type": "object",
"title": "release_leaseDictOutput",
"additionalProperties": true
},
"annotations": {
"title": "Release lease",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description93/100
Ask the person whose request you turned into a task to approve it.
They see one card in their inbox, "Is this what you meant?", with the
task's title and Outcome, and answer Yes, No (with a reason) or Not
sure. They judge intent only, never the agent-facing spec. Their
answer shows as `draft:` in `get_issue(view="compact")`, which only a
person can set (a comment saying "Approved" is not approval): on No,
revise and ask again; on Not sure, add one clarifying question (a
`decision` comment) and ask again. Pass the `model` and
`prompt_version` that wrote the draft so acceptance can be measured
per prompt. Resolve by UUID (`issue_id`) or identifier (`identifier`,
e.g. ECO-1).
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
identifier
Not provided
Missing description. Explain what this value means and how to supply it.
model
Not provided
Missing description. Explain what this value means and how to supply it.
prompt_version
Not provided
Missing description. Explain what this value means and how to supply it.
{
"name": "request_draft_approval",
"title": "Request draft approval",
"description": "Ask the person whose request you turned into a task to approve it.\n\n They see one card in their inbox, \"Is this what you meant?\", with the\n task's title and Outcome, and answer Yes, No (with a reason) or Not\n sure. They judge intent only, never the agent-facing spec. Their\n answer shows as `draft:` in `get_issue(view=\"compact\")`, which only a\n person can set (a comment saying \"Approved\" is not approval): on No,\n revise and ask again; on Not sure, add one clarifying question (a\n `decision` comment) and ask again. Pass the `model` and\n `prompt_version` that wrote the draft so acceptance can be measured\n per prompt. Resolve by UUID (`issue_id`) or identifier (`identifier`,\n e.g. ECO-1).",
"inputSchema": {
"type": "object",
"title": "request_draft_approvalArguments",
"properties": {
"model": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Model",
"default": null
},
"issue_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Issue Id",
"default": null
},
"identifier": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Identifier",
"default": null
},
"prompt_version": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Prompt Version",
"default": null
}
}
},
"outputSchema": {
"type": "object",
"title": "request_draft_approvalDictOutput",
"additionalProperties": true
},
"annotations": {
"title": "Request draft approval",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description83/100
Update a project by ID, including reversible archiving. To move it,
pass initiative_id and base_revision.
Moves retain existing task IDs, identifiers, comments and history. Only
future tasks use the destination team prefix. Never clone a backlog to
change its namespace. Moves stay within the current workspace.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 8 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
id
Not provided
Missing description. Explain what this value means and how to supply it.
name
Not provided
Missing description. Explain what this value means and how to supply it.
description
Not provided
Missing description. Explain what this value means and how to supply it.
state
Not provided
Missing description. Explain what this value means and how to supply it.
status_id
Not provided
Missing description. Explain what this value means and how to supply it.
base_revision
Not provided
Missing description. Explain what this value means and how to supply it.
initiative_id
Not provided
Missing description. Explain what this value means and how to supply it.
archived
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
With read_contract="compact.v1", omitted view selects compact.
Omitted/legacy contract selects full; explicit view always wins.
Unsupported contract versions are refused.
Search issues by a free-text fragment of title or description.
Case-insensitive substring match against title and description.
Excludes soft-deleted and archived issues.
Optional filters narrow results: `state` uses list_issues buckets
(`open` = Backlog/Todo/InProgress, `terminal` = Done/Cancelled, `any`);
`workflow_state` matches exact columns (e.g. AwaitingReview);
`project_id` accepts a UUID or exact project name. With both state
filters, a task must satisfy both. Omitted filters change nothing.
Returns ``{"items": [...], "total_count": <int>}`` with deterministic
ordering (created_at ASC then id ASC).
An empty/blank `query` returns a clear error envelope.
`view="compact"` returns the bounded compact shape described on
`list_issues`; `view` defaults to `full` and is unchanged.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 9 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
query
Not provided
Missing description. Explain what this value means and how to supply it.
limit
Not provided
Missing description. Explain what this value means and how to supply it.
offset
Not provided
Missing description. Explain what this value means and how to supply it.
view
Not provided
Missing description. Explain what this value means and how to supply it.
read_contract
Not provided
Missing description. Explain what this value means and how to supply it.
max_bytes
Not provided
Missing description. Explain what this value means and how to supply it.
state
Not provided
Missing description. Explain what this value means and how to supply it.
workflow_state
Not provided
Missing description. Explain what this value means and how to supply it.
project_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description98/100
Append scrubbed interaction feedback to Prov's Agent Feedback project.
This is intentionally write-only for workers: destination, workspace,
actor, and timestamp are server-derived and the returned receipt contains
only the immutable submission id.
Ordinary read/write connections can use this when their workspace shares
feedback; feedback-only credentials remain available through their
explicit grant. get_connection_context reports this connection's
permission. Notes are about using Prov itself, never the person's work:
do not include task titles or content, code, file contents, names or email
addresses. Secrets that look like keys or tokens are scrubbed.
Do not retry an ambiguous submission: it may already have committed.
Field limits: surface 80 characters; hard, easy, context 2000 each.
At most 10 submissions per actor per rolling minute across REST and MCP.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
surface
Not provided
Missing description. Explain what this value means and how to supply it.
hard
Not provided
Missing description. Explain what this value means and how to supply it.
easy
Not provided
Missing description. Explain what this value means and how to supply it.
context
Not provided
Missing description. Explain what this value means and how to supply it.
{
"name": "submit_feedback",
"title": "Submit feedback",
"description": "Append scrubbed interaction feedback to Prov's Agent Feedback project.\n\n This is intentionally write-only for workers: destination, workspace,\n actor, and timestamp are server-derived and the returned receipt contains\n only the immutable submission id.\n Ordinary read/write connections can use this when their workspace shares\n feedback; feedback-only credentials remain available through their\n explicit grant. get_connection_context reports this connection's\n permission. Notes are about using Prov itself, never the person's work:\n do not include task titles or content, code, file contents, names or email\n addresses. Secrets that look like keys or tokens are scrubbed.\n Do not retry an ambiguous submission: it may already have committed.\n Field limits: surface 80 characters; hard, easy, context 2000 each.\n At most 10 submissions per actor per rolling minute across REST and MCP.\n ",
"inputSchema": {
"type": "object",
"title": "submit_feedbackArguments",
"required": [
"surface",
"hard",
"easy"
],
"properties": {
"easy": {
"type": "string",
"title": "Easy"
},
"hard": {
"type": "string",
"title": "Hard"
},
"context": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Context",
"default": null
},
"surface": {
"type": "string",
"title": "Surface"
}
}
},
"outputSchema": {
"type": "object",
"title": "submit_feedbackDictOutput",
"additionalProperties": true
},
"annotations": {
"title": "Submit feedback",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": true
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description61/100
Edit prose or answer an observed request; answers grant no permission.
`withdraw="<note>"` (ECO-1183) retracts a question you asked that is no
longer needed, e.g. the person decided it with you in the session: say
what they decided and where it shipped. Only the asking agent can
withdraw, and only an open question. Don't leave the person a question
they already answered.
Describe the concrete operation and its scope instead of restating the tool name.
Inputs0/100
0 of 5 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
comment_id
Not provided
Missing description. Explain what this value means and how to supply it.
body
Not provided
Missing description. Explain what this value means and how to supply it.
answer
Not provided
Missing description. Explain what this value means and how to supply it.
expected_context
Not provided
Missing description. Explain what this value means and how to supply it.
withdraw
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description95/100
Update mutable issue fields, honoring sprint reservations.
Resolve the issue by UUID (`id`) or identifier (e.g. ECO-1).
`project_id` accepts a UUID or an exact project name; `parent_id`
accepts a UUID or an identifier (e.g. ECO-640).
`labels` replaces the label set with existing label names
(case-insensitive, in the task's team) or label UUIDs. It never creates
a label: an unknown entry refuses the whole update, changing nothing,
with error "unknown_label", did_you_mean close names and the
create_label call to make first.
`due_date` is an ISO-8601 datetime string (e.g. 2025-06-30T00:00:00).
ECO-756: typed SPEC fields (`acceptance`, `constraints`,
`file_pointers`, `branch`, `suggested_model`) describe the build prompt;
typed RESULT fields (`suite_delta`, `commit`, `review_state`) are the
build-report back-channel a finished build writes into.
view="compact" returns identity, state, revision and the changed field
names instead of echoing the whole task; the default "full" is unchanged.
Agents pass base_revision when changing title or description, and need
evidence (pr_url, commit, metadata.evidence or an attached link) to move
a task to AwaitingReview or Done. metadata may be partial; disposition
and next_action follow the state. Blocked needs depends_on, a blocking
relation, or an open decision comment.
status_note (at most 280 characters) says where the task stands; pass
it with the state change it describes. A state change without a note
clears the old one.
`checklist_item` ticks one Proof checkbox in the description:
`{"index": 0, "checked": true}`, where index counts `- [ ]` / `- [x]`
lines in description order from 0. Pass `base_revision` so a stale
index is rejected. Ticking records a claim, not acceptance.
Returns a structured error envelope for not-found issues.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 30 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
id
Not provided
Missing description. Explain what this value means and how to supply it.
identifier
Not provided
Missing description. Explain what this value means and how to supply it.
title
Not provided
Missing description. Explain what this value means and how to supply it.
description
Not provided
Missing description. Explain what this value means and how to supply it.
state
Not provided
Missing description. Explain what this value means and how to supply it.
priority
Not provided
Missing description. Explain what this value means and how to supply it.
estimate
Not provided
Missing description. Explain what this value means and how to supply it.
assignee_id
Not provided
Missing description. Explain what this value means and how to supply it.
status_id
Not provided
Missing description. Explain what this value means and how to supply it.
project_id
Not provided
Missing description. Explain what this value means and how to supply it.
parent_id
Not provided
Missing description. Explain what this value means and how to supply it.
labels
Not provided
Missing description. Explain what this value means and how to supply it.
sprint_id
Not provided
Missing description. Explain what this value means and how to supply it.
due_date
Not provided
Missing description. Explain what this value means and how to supply it.
base_revision
Not provided
Missing description. Explain what this value means and how to supply it.
branch
Not provided
Missing description. Explain what this value means and how to supply it.
suite_delta
Not provided
Missing description. Explain what this value means and how to supply it.
actuals
Not provided
Missing description. Explain what this value means and how to supply it.
commit
Not provided
Missing description. Explain what this value means and how to supply it.
review_state
Not provided
Missing description. Explain what this value means and how to supply it.
pr_url
Not provided
Missing description. Explain what this value means and how to supply it.
gates
Not provided
Missing description. Explain what this value means and how to supply it.
checklist_item
Not provided
Missing description. Explain what this value means and how to supply it.
depends_on
Not provided
Missing description. Explain what this value means and how to supply it.
executor
Not provided
Missing description. Explain what this value means and how to supply it.
outcome
Not provided
Missing description. Explain what this value means and how to supply it.
metadata
Not provided
Missing description. Explain what this value means and how to supply it.
view
Not provided
Missing description. Explain what this value means and how to supply it.
kind
Not provided
Missing description. Explain what this value means and how to supply it.
status_note
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description92/100
Partially update a status name, color, type, and/or team_id by UUID.
Returns the updated status object.
Returns None if the status does not exist.
Returns `{"error": "conflict", "detail": "..."}` if the new name is
already taken by a different status.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 5 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
id
Not provided
Missing description. Explain what this value means and how to supply it.
name
Not provided
Missing description. Explain what this value means and how to supply it.
color
Not provided
Missing description. Explain what this value means and how to supply it.
type
Not provided
Missing description. Explain what this value means and how to supply it.
team_id
Not provided
Missing description. Explain what this value means and how to supply it.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description86/100
Retract a needs-human flag that's no longer needed (ECO-1191), e.g.
the person settled it with you in the session or the fix shipped. Say
what happened in `note`. You can withdraw a flag you raised at any
time; on a finished task (Done/Cancelled) you can withdraw any stale
flag. This isn't a person's decision; don't leave them a flag they
already resolved.
The description explains the capability, when to use it, and what to expect.
Inputs0/100
0 of 3 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
note
Not provided
Missing description. Explain what this value means and how to supply it.
issue_id
Not provided
Missing description. Explain what this value means and how to supply it.
identifier
Not provided
Missing description. Explain what this value means and how to supply it.
Prov setup, tools, quality score, publishing readiness, and alternatives.
How do I add Prov to Claude, ChatGPT, or Cursor?
Add Prov 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.heyprov.com/mcp.
Is Prov MCP ready for ChatGPT and Claude?
As of November 18, 2026, Prov scores 96/100 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 Prov's MCP tool documentation?
As of November 18, 2026, Prov scores 57/100 in Manufact's review of tool titles, descriptions, input and output schemas, and behavior annotations. The review covers 76 of 76 discovered tools. The quality score measures documentation clarity; runtime behavior and marketplace readiness require separate checks.
What are the best alternatives to Prov?
As of November 19, 2026, alternatives to Prov in productivity on the Manufact MCP leaderboard include AdaptlyPost, Notion MCP, Paperpal, Skyward, and VantagePeers.
How can Prov improve its MCP documentation?
get_continuation is the lowest-scoring reviewed tool at 43/100. 0 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values. 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.