--- name: agentspec-one.vercel.app description: agentspec-one.vercel.app provides static analysis and metadata generation tools for API and MCP service definitions. It scores OpenAPI and MCP specs for agent-readiness, identifies risky or under-specified patterns, and generates structured x402 Bazaar listing metadata from those reports. host: agentspec-one.vercel.app --- # agentspec-one.vercel.app This host serves developers and agents preparing APIs or MCP servers for agent-marketplace publication. It covers two distinct formats (OpenAPI and MCP tool manifests) and two distinct tasks: readiness evaluation (scoring, grading, risk flagging) and metadata generation (producing publishable x402 Bazaar drafts). It does not perform runtime monitoring, syntax validation, or full documentation generation. ## When to use this host Use this host when you need to evaluate an OpenAPI spec or MCP tool manifest for agent-readiness, or when you need to generate a structured x402 Bazaar listing draft from an existing report. For a quick numeric gate before committing to full analysis, use score-agent-readiness first. For non-OpenAPI formats other than MCP (e.g., GraphQL, gRPC, plain REST without an OpenAPI spec), this host has no applicable skill. For runtime health checks, uptime monitoring, or live API testing, use a dedicated monitoring or testing service instead. For generating full OpenAPI specs or service documentation from scratch, use a documentation-generation tool rather than this host. ## Capabilities ### Spec Inspection and Scoring Evaluates OpenAPI or MCP service definitions for agent-readiness, returning numeric scores, letter grades, per-endpoint or per-tool risk analysis, and marketplace candidacy signals. - **`inspect-openapi-spec`** — Accepts an OpenAPI spec object and returns an agent-readiness score, grade, per-endpoint risk analysis, recommendations, and x402 Bazaar candidates. - **`inspect-mcp-tools`** — Analyzes an MCP server's tool definitions and returns an agent-readiness score, grade, per-tool risk levels, warnings, and marketplace candidates. - **`score-agent-readiness`** — Scores an OpenAPI or MCP spec for agent-readiness, returning a numeric score, letter grade, summary, risk count, and recommendation count. ### Bazaar Metadata Generation Transforms an AgentSpec report or service description into a structured x402 Bazaar metadata draft, including name, category, tags, pricing suggestion, input schema, and output example. - **`generate-bazaar-metadata`** — Accepts an AgentSpec report or service description and returns a structured x402 Bazaar metadata draft including name, category, tags, price suggestion, input schema, output example, and readiness notes. ## Workflows ### OpenAPI to Bazaar Listing *Use when an agent needs to take a raw OpenAPI spec and produce a publishable x402 Bazaar metadata draft in one pipeline.* 1. **`inspect-openapi-spec`** — Analyze the OpenAPI spec to produce an agent-readiness report including score, grade, per-endpoint risk analysis, and Bazaar candidates. 2. **`generate-bazaar-metadata`** — Feed the inspection report into the metadata generator to produce a structured x402 Bazaar listing draft with pricing guidance and discoverability fields. ### MCP Server to Bazaar Listing *Use when an agent needs to evaluate an MCP server's tools and produce a publishable x402 Bazaar metadata draft from the results.* 1. **`inspect-mcp-tools`** — Analyze the MCP tool manifest to produce an agent-readiness report including score, grade, per-tool risk levels, and marketplace candidates. 2. **`generate-bazaar-metadata`** — Feed the MCP inspection report into the metadata generator to produce a structured x402 Bazaar listing draft. ### Quick Readiness Gate *Use when an agent needs a fast pass/fail signal on either an OpenAPI or MCP spec before deciding whether to invest in full inspection or listing.* 1. **`score-agent-readiness`** — Submit the spec (OpenAPI or MCP) to get a numeric score, letter grade, risk count, and recommendation count as a lightweight pre-filter. 2. **`inspect-openapi-spec`** — If the score warrants deeper analysis, run the full OpenAPI inspector to get per-endpoint detail and Bazaar candidates. ## Skill reference ### `generate-bazaar-metadata` **Bazaar Metadata Generator** — Accepts an AgentSpec report or service description and returns a structured x402 Bazaar metadata draft including name, category, tags, price suggestion, input schema, output example, and readiness notes. *Use when:* Use when an agent or developer has an AgentSpec report or service description and needs to produce a publishable x402 Bazaar metadata draft for a specific endpoint, including pricing guidance and discoverability fields. *Not for:* Do not use for grading or scoring an API — use an AgentSpec inspection endpoint instead. Not suitable for generating OpenAPI specs or full service documentation. **Inputs:** - `report` (object, required) — An AgentSpec report object describing the service, including name, type, grade, score, and bazaarCandidates array with endpoint details such as path, method, description, inputSchema, and outputExample. **Returns:** Returns a Bazaar metadata draft with name, description, category, tags, method, path, a priceSuggestion object, a nested bazaarDraft preserving inputSchema and outputExample, and an array of readinessNotes. **Example:** `{"report":{"name":"Weather Utility API","type":"openapi","grade":"A","score":100,"bazaarCandidates":[{"name":"getForecast","path":"/forecast","method":"GET","description":"Get a weather forecast by city.","inputSchema":{"type":"object","required":["city"],"properties":{"city":{"type":"string"}}},"outputExample":{"data":{"city":"Paris"},"status":200}}]}}` --- ### `inspect-openapi-spec` **OpenAPI Agent Inspector** — Accepts an OpenAPI spec object and returns an agent-readiness score, grade, per-endpoint risk analysis, recommendations, and x402 Bazaar candidates. *Use when:* Use when an agent or developer needs to evaluate an OpenAPI spec for agent-readiness, identify missing descriptions or risky patterns, and generate x402 Bazaar listing candidates before publishing an API. *Not for:* Do not use for validating OpenAPI syntax or schema correctness alone; this service scores agent-readiness and monetization fit, not spec validity. Not suitable for non-OpenAPI API formats. **Inputs:** - `spec` (object, required) — A valid OpenAPI 3.x spec object (JSON). Must include at minimum `openapi`, `info`, and `paths` fields. Additional properties are accepted. **Returns:** Returns a score (0-100), letter grade, per-endpoint risk and agentScore breakdown, spec-level recommendations, and a bazaarCandidates array with price suggestions for each publishable endpoint. **Example:** `{"spec":{"openapi":"3.1.0","info":{"title":"Weather Utility API","version":"1.0.0"},"paths":{"/forecast":{"get":{"summary":"Get a weather forecast by city.","operationId":"getForecast","responses":{"200":{"description":"Forecast result.","content":{"application/json":{"schema":{"type":"object"},"example":{"city":"Paris","days":[]}}}}}}}}}}` --- ### `inspect-mcp-tools` **MCP Inspector** — Analyzes an MCP server's tool definitions and returns an agent-readiness score, grade, per-tool risk levels, warnings, and marketplace candidates. *Use when:* Use when an agent or developer needs to evaluate an MCP server's tools for launch readiness, identify risky or under-specified tools, and discover which tools are suitable for marketplace listing. *Not for:* Do not use for inspecting non-MCP APIs or OpenAPI specs; this endpoint expects an MCP tool manifest with a tools array. Not a runtime monitoring tool — it performs static analysis only. **Inputs:** - `name` (string) — Human-readable name of the MCP service being inspected. - `tools` (array) — Array of MCP tool objects, each with a name, description, and inputSchema, to be analyzed for readiness and risk. **Returns:** Returns score=96, grade=A, one tool entry with riskLevel=low and agentScore=96, empty risks/warnings/recommendations, and one bazaarCandidate with a suggested price of 0.03 USDC per call. **Example:** `{"name": "Repo Assistant MCP", "tools": [{"name": "inspect_repository", "description": "Analyze a public GitHub repository and return detected stack.", "inputSchema": {"type": "object", "required": ["repoUrl"], "properties": {"repoUrl": {"type": "string"}}, "additionalProperties": false}}]}` --- ### `score-agent-readiness` **Agent Readiness Score** — Scores an OpenAPI or MCP spec for agent-readiness, returning a numeric score, letter grade, summary, risk count, and recommendation count. *Use when:* Use when an agent or developer needs to evaluate how ready an OpenAPI or MCP service definition is for agent consumption, x402 monetization, or listing on an agent marketplace. *Not for:* Do not use for runtime API health checks or uptime monitoring; this scores a spec document, not a live service. **Inputs:** - `spec` (object, required) — The OpenAPI or MCP spec object to evaluate. Pass the full spec as a nested JSON object under this key. **Returns:** Returns type='openapi', score=93, grade='A', a plain-English summary of endpoint quality and risk signals, riskCount=0, and recommendationCount=2. **Example:** `{"spec":{"info":{"title":"Mini API","version":"1.0.0"},"paths":{"/ping":{"get":{"summary":"Ping service","operationId":"ping","responses":{"200":{"content":{"application/json":{"schema":{"type":"object"},"example":{"ok":true}}},"description":"pong"}}}}},"openapi":"3.1.0"}}` ---