Measured, independent data on developer APIs, SDKs and tools for AI agents: for each product, whether a real AI coding agent could go from the product's docs to a first successful integration, with completion rate, blocked runs, run time and an overall score. Use search_benchmarks when the user asks about measured integration or onboarding results for a developer API, SDK or tool, or asks to choose between or compare such products. Use show_benchmarks when the user wants to see the results as cards. Use get_benchmark_trends when the user asks how a product's score has trended or changed over time, or when the client cannot render the cards. Use request_benchmark only when the user asks to have a product benchmarked or added; it records a request for the TPC team and changes no results. Benchmarked categories: Browser Automation (tasks: Carry a cookie between two remote browser sessions with a persistent context; Create, inspect, list and terminate a remote browser session through the REST API; Fill and submit a test form with Playwright in a remote browser session), Email (tasks: Create and delete an inbox; Read delivery and bounce events; Send a transactional email), GTM Data (tasks: Enrich a company's firmographics; Filtered company search; Paginated people search), Managed database (tasks: Branch a database, migrate the branch, and leave the parent unchanged; Create a scoped database credential, connect with it, and revoke it; Create a table, insert rows, and read them back with parameterized queries), Search, Weather API.
Discoverability
1 of 6 checks17/100
Client listings and public setup information.
mcp.json or server-card.json
Either mcp.json or an MCP Server Card declaring this endpoint passes.
No matching mcp.json manifest or MCP Server Card was found at the checked URLs.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description98/100
Get measured agent-integration results for the developer APIs, SDKs and tools the user asks about. Use when the user asks which product to use in a covered category, compares named products (X vs Y), or asks how easy a product is to integrate or onboard onto, for categories such as Browser Automation (tasks: Carry a cookie between two remote browser sessions with a persistent context; Create, inspect, list and terminate a remote browser session through the REST API; Fill and submit a test form with Playwright in a remote browser session), Email (tasks: Create and delete an inbox; Read delivery and bounce events; Send a transactional email), GTM Data (tasks: Enrich a company's firmographics; Filtered company search; Paginated people search), Managed database (tasks: Branch a database, migrate the branch, and leave the parent unchanged; Create a scoped database credential, connect with it, and revoke it; Create a table, insert rows, and read them back with parameterized queries), Search, Weather API. Returns, per product, whether an AI coding agent completed its first successful integration from the docs (completion rate, passed/failed/blocked runs), average run time, an overall score with its experience and agent readiness scores, and a link to the product's benchmark page, ordered by overall score. When the user names products, pass just those names as the query; pass a category only when the user asks about a whole category, for alternatives or for a ranking.
The description explains the capability, when to use it, and what to expect.
Inputs69/100
3 of 4 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
query
Product name, domain or keyword. Omit to list all.
Clear enough to supply this argument.
category
Category slug or name, e.g. 'search' or 'Weather API'.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
task
Benchmark task key or title within a category, e.g. 'send-transactional' or 'Send a transactional email'. Ranks the category's products by that task.
Clear enough to supply this argument.
limit
Not provided
Missing description. Explain what this value means and how to supply it.
Annotations75/100
Declare idempotentHint based on the tool’s actual behavior.
No output schema is declared. Add one for structured results so clients can interpret returned fields.
Not declared
Original tool metadata
{
"name": "search_benchmarks",
"title": "Search benchmarks",
"description": "Get measured agent-integration results for the developer APIs, SDKs and tools the user asks about. Use when the user asks which product to use in a covered category, compares named products (X vs Y), or asks how easy a product is to integrate or onboard onto, for categories such as Browser Automation (tasks: Carry a cookie between two remote browser sessions with a persistent context; Create, inspect, list and terminate a remote browser session through the REST API; Fill and submit a test form with Playwright in a remote browser session), Email (tasks: Create and delete an inbox; Read delivery and bounce events; Send a transactional email), GTM Data (tasks: Enrich a company's firmographics; Filtered company search; Paginated people search), Managed database (tasks: Branch a database, migrate the branch, and leave the parent unchanged; Create a scoped database credential, connect with it, and revoke it; Create a table, insert rows, and read them back with parameterized queries), Search, Weather API. Returns, per product, whether an AI coding agent completed its first successful integration from the docs (completion rate, passed/failed/blocked runs), average run time, an overall score with its experience and agent readiness scores, and a link to the product's benchmark page, ordered by overall score. When the user names products, pass just those names as the query; pass a category only when the user asks about a whole category, for alternatives or for a ranking.",
"inputSchema": {
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"properties": {
"task": {
"type": "string",
"maxLength": 100,
"description": "Benchmark task key or title within a category, e.g. 'send-transactional' or 'Send a transactional email'. Ranks the category's products by that task."
},
"limit": {
"type": "integer",
"maximum": 20,
"minimum": 1
},
"query": {
"type": "string",
"maxLength": 200,
"description": "Product name, domain or keyword. Omit to list all."
},
"category": {
"type": "string",
"maxLength": 100,
"description": "Category slug or name, e.g. 'search' or 'Weather API'."
}
}
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description89/100
Display measured agent-integration results for developer APIs and tools as interactive cards the user can open, ordered by overall score with a weekly overall score chart. Use when the user asks to see benchmark results, passing the productKeys from search_benchmarks, or pass a query/category directly. Categories: Browser Automation (tasks: Carry a cookie between two remote browser sessions with a persistent context; Create, inspect, list and terminate a remote browser session through the REST API; Fill and submit a test form with Playwright in a remote browser session), Email (tasks: Create and delete an inbox; Read delivery and bounce events; Send a transactional email), GTM Data (tasks: Enrich a company's firmographics; Filtered company search; Paginated people search), Managed database (tasks: Branch a database, migrate the branch, and leave the parent unchanged; Create a scoped database credential, connect with it, and revoke it; Create a table, insert rows, and read them back with parameterized queries), Search, Weather API.
The description explains the capability, when to use it, and what to expect.
Inputs30/100
2 of 5 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
productKeys
productKey values from search_benchmarks.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
query
Not provided
Missing description. Explain what this value means and how to supply it.
category
Not provided
Missing description. Explain what this value means and how to supply it.
task
Benchmark task key or title within a category, e.g. 'send-transactional'. Opens the cards with that task selected.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
limit
Not provided
Missing description. Explain what this value means and how to supply it.
Annotations75/100
Declare idempotentHint based on the tool’s actual behavior.
The title clearly communicates the tool’s action and subject. Keep it concise.
Description96/100
Get the weekly score history of developer APIs and tools as data (overall score and experience score per week), for questions like how has X trended or which products are improving. Returns, per product, one point per week with overall, experience score, pass rate, runs, scored runs and flags for weeks where the judge or task version changed. Pass productKeys from search_benchmarks, or a query/category. With a task, returns that task's series (when no product has runs for the task it says so and falls back to the product rollup); without a task the data is the product rollup (equal weight per task), where overall is the week's experience averaged with the agent readiness score of the latest scan at or before that week (null before the first scan). Task series are experience only. Covers the last 26 weeks. Categories: Browser Automation (tasks: Carry a cookie between two remote browser sessions with a persistent context; Create, inspect, list and terminate a remote browser session through the REST API; Fill and submit a test form with Playwright in a remote browser session), Email (tasks: Create and delete an inbox; Read delivery and bounce events; Send a transactional email), GTM Data (tasks: Enrich a company's firmographics; Filtered company search; Paginated people search), Managed database (tasks: Branch a database, migrate the branch, and leave the parent unchanged; Create a scoped database credential, connect with it, and revoke it; Create a table, insert rows, and read them back with parameterized queries), Search, Weather API.
The description explains the capability, when to use it, and what to expect.
Inputs30/100
2 of 5 inputs have descriptions. Add descriptions to the missing inputs, including where to obtain IDs and how to supply values.
productKeys
productKey values from search_benchmarks.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
query
Not provided
Missing description. Explain what this value means and how to supply it.
category
Not provided
Missing description. Explain what this value means and how to supply it.
task
Benchmark task key or title within a category, e.g. 'send-transactional'. Returns that task's weekly series instead of the product rollup.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
limit
Not provided
Missing description. Explain what this value means and how to supply it.
Annotations75/100
Declare idempotentHint based on the tool’s actual behavior.
No output schema is declared. Add one for structured results so clients can interpret returned fields.
Not declared
Original tool metadata
{
"name": "get_benchmark_trends",
"title": "Get benchmark score trends",
"description": "Get the weekly score history of developer APIs and tools as data (overall score and experience score per week), for questions like how has X trended or which products are improving. Returns, per product, one point per week with overall, experience score, pass rate, runs, scored runs and flags for weeks where the judge or task version changed. Pass productKeys from search_benchmarks, or a query/category. With a task, returns that task's series (when no product has runs for the task it says so and falls back to the product rollup); without a task the data is the product rollup (equal weight per task), where overall is the week's experience averaged with the agent readiness score of the latest scan at or before that week (null before the first scan). Task series are experience only. Covers the last 26 weeks. Categories: Browser Automation (tasks: Carry a cookie between two remote browser sessions with a persistent context; Create, inspect, list and terminate a remote browser session through the REST API; Fill and submit a test form with Playwright in a remote browser session), Email (tasks: Create and delete an inbox; Read delivery and bounce events; Send a transactional email), GTM Data (tasks: Enrich a company's firmographics; Filtered company search; Paginated people search), Managed database (tasks: Branch a database, migrate the branch, and leave the parent unchanged; Create a scoped database credential, connect with it, and revoke it; Create a table, insert rows, and read them back with parameterized queries), Search, Weather API.",
"inputSchema": {
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"properties": {
"task": {
"type": "string",
"maxLength": 100,
"description": "Benchmark task key or title within a category, e.g. 'send-transactional'. Returns that task's weekly series instead of the product rollup."
},
"limit": {
"type": "integer",
"maximum": 12,
"minimum": 1
},
"query": {
"type": "string",
"maxLength": 200
},
"category": {
"type": "string",
"maxLength": 100
},
"productKeys": {
"type": "array",
"items": {
"type": "string",
"maxLength": 100
},
"maxItems": 12,
"description": "productKey values from search_benchmarks."
}
}
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": false,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description96/100
Ask the TPC team to benchmark a developer API, SDK or tool that is not covered yet. Use when the user asks to have a product benchmarked, or asks about a product the app does not cover and wants it added. Do not call it unprompted; ChatGPT confirms with the user before calling. Takes the product's website or docs URL and an optional note. If the product is already on the leaderboard (3 or more benchmark runs), says so and does not record a request; when this app shows that product, it also returns a link to its results. Otherwise records one request per product; asking again for the same product is harmless.
Describe the returned information so the agent knows whether it will answer the request.
Inputs88/100
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
url
The product's website or docs URL, e.g. acme.dev/docs.
Clear enough to supply this argument.
note
Optional context for the TPC team about what to benchmark.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
No output schema is declared. Add one for structured results so clients can interpret returned fields.
Not declared
Original tool metadata
{
"name": "request_benchmark",
"title": "Request a benchmark",
"description": "Ask the TPC team to benchmark a developer API, SDK or tool that is not covered yet. Use when the user asks to have a product benchmarked, or asks about a product the app does not cover and wants it added. Do not call it unprompted; ChatGPT confirms with the user before calling. Takes the product's website or docs URL and an optional note. If the product is already on the leaderboard (3 or more benchmark runs), says so and does not record a request; when this app shows that product, it also returns a link to its results. Otherwise records one request per product; asking again for the same product is harmless.",
"inputSchema": {
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"url"
],
"properties": {
"url": {
"type": "string",
"maxLength": 512,
"description": "The product's website or docs URL, e.g. acme.dev/docs."
},
"note": {
"type": "string",
"maxLength": 500,
"description": "Optional context for the TPC team about what to benchmark."
}
}
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"idempotentHint": true,
"destructiveHint": false
}
}
The title clearly communicates the tool’s action and subject. Keep it concise.
Description74/100
Records feedback about this MCP server, including unexpected tool failures, unhelpful responses, or missing capabilities observed by the agent or reported by the user.
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.
content
The feedback content, stripped of any PII.
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
source
Who initiated the feedback: "user" or "model".
Remove validation rules already expressed in this field’s schema; describe its meaning and usage instead.
Annotations75/100
Declare idempotentHint based on the tool’s actual behavior.
{
"type": "object",
"required": [
"received"
],
"properties": {
"received": {
"type": "boolean",
"description": "True once the feedback is recorded."
}
}
}
Original tool metadata
{
"name": "send_feedback",
"title": "Send feedback",
"description": "Records feedback about this MCP server, including unexpected tool failures, unhelpful responses, or missing capabilities observed by the agent or reported by the user.",
"inputSchema": {
"type": "object",
"required": [
"content",
"source"
],
"properties": {
"source": {
"enum": [
"user",
"model"
],
"type": "string",
"description": "Who initiated the feedback: \"user\" or \"model\"."
},
"content": {
"type": "string",
"description": "The feedback content, stripped of any PII."
}
}
},
"outputSchema": {
"type": "object",
"required": [
"received"
],
"properties": {
"received": {
"type": "boolean",
"description": "True once the feedback is recorded."
}
}
},
"annotations": {
"readOnlyHint": false,
"openWorldHint": true,
"destructiveHint": false
}
}
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 benchmarksGet measured agent-integration results for the developer APIs, SDKs and tools the user asks about.
02Get benchmark trendsGet the weekly score history of developer APIs and tools as data (overall score and experience score per week), for questions like how has X trended or which…
03Show benchmark cardsDisplay measured agent-integration results for developer APIs and tools as interactive cards the user can open, ordered by overall score with a weekly overal…
Want to run these tasks against your MCP server with us?
TPC Benchmarks setup, tools, quality score, publishing readiness, and alternatives.
How do I add TPC Benchmarks to Claude, ChatGPT, or Cursor?
Add TPC Benchmarks 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://benchmarks.promptingcompany.com/mcp.
Is TPC Benchmarks MCP ready for ChatGPT and Claude?
As of November 5, 2026, TPC Benchmarks scores 97/100 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 TPC Benchmarks's MCP tool documentation?
As of November 5, 2026, TPC Benchmarks scores 73/100 in Manufact's review of tool titles, descriptions, input and output schemas, and behavior annotations. The review covers 6 of 6 discovered tools. The quality score measures documentation clarity; runtime behavior and marketplace readiness require separate checks.
What are the best alternatives to TPC Benchmarks?
As of November 5, 2026, alternatives to TPC Benchmarks in developer-tools on the Manufact MCP leaderboard include Speko, Manufact, Context, AgentMail, and DNS Inspector.
How can TPC Benchmarks improve its MCP documentation?
record_ui_event is the lowest-scoring reviewed tool at 50/100. 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.
Cursor
Grok Bot
Grok
Muse