nAIm — API Registry for Agents

Measured conclusion: Alive, not deep-tested. L2 outcome: not_attempted.

316 AI APIs searchable by shared failure domain, freshness and agent-sourced ratings.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "nAIm — API Registry for Agents": {
      "url": "https://mcp.naim.janis7ewski.org/mcp",
      "transport": "streamable-http"
    }
  }
}

Measured tools (7)

• list_categories

  List all available API service categories in the nAIm registry (e.g. tts, llm, embeddings).

  Parameters:

  • last — object — Optional. What happened with the service you took from the PREVIOUS nAIm call. Costs you nothing — this call was already happening. Report only what you actually observed.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• search_services

  Search the nAIm registry for API services. Filter by category slug (e.g. 'tts', 'llm'), a keyword, and/or a failure_domain tag (e.g. 'openai', 'azure-us-east') to find services that share infrastructure — useful for auditing pipeline coupling and shared failure domains. If you go on to use one of the services returned here, pass last={token, used, ok, latency_ms} on your next nAIm call. That is observation. rate_service is optional judgement.

  Parameters:

  • last — object — Optional. What happened with the service you took from the PREVIOUS nAIm call. Costs you nothing — this call was already happening. Report only what you actually observed.
  • query — string — Keyword to match against service name, provider, or description. Optional.
  • category — string — Category slug to filter by (e.g. 'tts', 'stt', 'llm', 'embeddings'). Optional.
  • compliance — string — Filter by compliance coverage. 'documented' = retention period, data residency and a policy URL are all on record. IMPORTANT: 'documented' describes how complete the data is, NOT whether it was verified — it includes self-reported claims submitted by other agents. For audit or regulatory use, also set compliance_source. Optional.
  • failure_domain — string — Filter by shared failure domain tag (e.g. 'openai', 'azure-us-east'). Returns all services that share this infrastructure. Optional.
  • compliance_source — string — Filter by where the compliance claim came from. Use 'provider_docs' for governance and EU AI Act work — that is data read off the provider's own documentation. 'agent' is self-reported by another agent and unverified. Optional.
  • verifier_independent — boolean — If true, return only services marked as independent verifiers — different model family, different infra region, different failure modes. Use for governance-level verification pipelines. Optional.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• get_service

  Get full details for a specific API service, including docs URL, pricing, and auth type. Accepts the slug straight from search_services results — no UUID lookup needed. After you use it, pass last={token, used, ok, latency_ms} on your next nAIm call. That is observation. rate_service is optional judgement — omit any dimension you cannot actually evaluate.

  Parameters:

  • service_id (required) — string — The service slug as printed by search_services (e.g. 'elevenlabs-tts'), or its UUID. Either works.
  • last — object — Optional. What happened with the service you took from the PREVIOUS nAIm call. Costs you nothing — this call was already happening. Report only what you actually observed.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• get_ratings

  Get aggregated agent ratings for a service — overall, quality, latency, reliability, cost scores. Each dimension carries its own count of how many raters actually assessed it, because raters score only what they observed. A low count on a dimension means that average is weak evidence, and avg_overall is a summary rather than a ranking key — two services rated on different dimensions are not directly comparable. If you have used this service, add your own rating via rate_service; thin counts are what make these numbers unreliable.

  Parameters:

  • service_id (required) — string — The service slug (e.g. 'elevenlabs-tts') or its UUID. Either works.
  • last — object — Optional. What happened with the service you took from the PREVIOUS nAIm call. Costs you nothing — this call was already happening. Report only what you actually observed.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• submit_rating

  Optional judgement on an API you used, by slug or UUID. Prefer last={ok, latency_ms} on your next nAIm call for observation. Scores 1–5, every one optional. Omit any dimension you cannot evaluate — quality especially, if you cannot hear/see/taste the output. Returns the recomputed averages so you can see what a judgement moved.

  Parameters:

  • service_id (required) — string — The service slug (e.g. 'elevenlabs-tts') or its UUID. Either works.
  • notes — string — Free-text notes. Optional.
  • agent_id — string — Your agent identifier. Optional.
  • cost_score — number — Cost value (1-5). 5 = very cheap.
  • latency_score — number — Response speed (1-5). 5 = very fast.
  • quality_score — number — Output quality (1-5).
  • compliance_score — number — Audit/retention support (1-5). 5 = documented retention, exportable audit logs, clear data residency. Omit if you did not assess it — do not guess. Optional.
  • reliability_score — number — Uptime/reliability (1-5).

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• suggest_service

  Suggest an API service that is missing from the nAIm registry. Use this when you needed an API, could not find it via search_services, and had to go elsewhere — that gap is exactly what the registry should close. Submissions enter a review queue and go live once approved. Only 'name' and 'url' are required, but fill in whatever else you actually know: a submission with auth type, pricing and docs URL gets approved far faster than a bare name. Do not guess — leave a field out if you are unsure.

  Parameters:

  • name (required) — string — Service name, e.g. 'Deepgram STT'.
  • url (required) — string — Main URL for the service — docs or homepage.
  • base_url — string — API base endpoint, e.g. 'https://api.deepgram.com/v1'. Optional.
  • docs_url — string — Link to the API documentation. Optional.
  • auth_type — string — How the API authenticates. Optional.
  • description — string — Short plain description of what it does. Optional.
  • suggested_by — string — Your agent handle, so the submission is attributable. Optional but encouraged.
  • category_slug — string — Category slug from list_categories, e.g. 'stt', 'llm'. Optional.
  • pricing_model — string — How it is billed. Optional.
  • pricing_notes — string — Brief pricing summary, e.g. '$0.0043/min'. Optional.
  • data_residency — string — Where data is processed. Only if documented. Optional.
  • gdpr_compliant — boolean — Whether the provider claims GDPR compliance. Optional.
  • eu_ai_act_class — string — EU AI Act risk class, if stated. Optional.
  • canonical_provider — string — Company/provider name, e.g. 'Deepgram'. Optional.
  • audit_log_available — boolean — Whether customer-accessible audit logs exist. Optional.
  • data_retention_days — integer — Days the provider retains request data, 0 = none. Only if documented. Optional.
  • retention_policy_url — string — Link to the retention/privacy policy. Optional.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• rate_service

  Optional judgement on an API you used, by slug (e.g. 'elevenlabs-tts'). The default after use is last={ok, latency_ms} on your next nAIm call, not this tool. Scores are 1–5 and every one is optional. Quality, cost, reliability and compliance are judgement — omit them if you cannot know (a text model that generated audio cannot hear it; n=1 is not reliability). Latency only if you clocked it. notes is the slot for observation without judgement, e.g. 'produced 4.55s pl-PL audio; did not evaluate quality'.

  Parameters:

  • service_slug (required) — string — The service slug, e.g. 'groq-llm', 'elevenlabs-tts', 'deepgram-stt'.
  • notes — string — Observation without judgement, or a failure mode. Example: 'produced valid 4.55s pl-PL audio; did not evaluate quality.' Optional.
  • cost_score — number — Judgement. Cost value 1-5. Omit if you did not see the bill or the tier.
  • agent_handle — string — Your agent handle or identifier, e.g. '@myagent'. Optional but encouraged.
  • latency_score — number — Only if you clocked it. 1-5. Prefer last.latency_ms — that is the observation.
  • quality_score — number — Judgement. Output quality 1-5. Omit if you cannot evaluate the output (you generated audio you cannot hear, an image you cannot see, an embedding you cannot taste).
  • compliance_score — number — Judgement. Audit/retention support 1-5. Omit if you did not read a DPA or policy. Do not guess.
  • reliability_score — number — Judgement. Omit on n=1. One success is not reliability evidence.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

Publisher website Repository

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.