API Evangelist enrichment overlay — HAMi Website Discovery API

overlay: 1.0.0
info:
  title: API Evangelist enrichment overlay — HAMi Website Discovery API
  version: 1.0.0
x-provenance:
  generated: '2026-09-12'
  method: generated
  extends: openapi/hami-website-discovery-openapi.json
  source: >-
    Enhancements derived from artifacts in this repository — well-known/hami-well-known.yml,
    skills/_index.yml, conformance/hami-conformance.yml, authentication/hami-authentication.yml and
    conventions/hami-conventions.yml. The original spec at
    https://project-hami.io/.well-known/openapi.json is never mutated; everything here is additive.
  note: >-
    HAMi's own spec is deliberately minimal — four paths, 200-only, no tags, no operationIds. This
    overlay adds operationIds, tags, and the observed media types so the document is usable by
    tooling that keys on them, and annotates each path with what the document behind it actually is.
actions:
  - target: $.info
    update:
      x-api-evangelist-profile: https://apis.io/provider/hami
      x-contact-note: >-
        Discovery surface of the HAMi documentation website, not the HAMi WebUI API. The HAMi WebUI
        API is a separate, self-hosted gRPC/grpc-gateway contract; see grpc/ in this repository.
  - target: $.paths['/robots.txt'].get
    update:
      operationId: getRobotsTxt
      tags: [discovery]
      x-content-signal: 'ai-train=no, search=yes, ai-input=yes'
      x-content-signal-note: >-
        Observed 2026-09-12. Repeated for ten named AI crawler user-agents (GPTBot, OAI-SearchBot,
        Claude-Web, anthropic-ai, Google-Extended, Amazonbot, Bytespider, CCBot, Applebot-Extended)
        and the wildcard agent. /search and any query-string URL are disallowed.
  - target: $.paths['/.well-known/api-catalog'].get
    update:
      operationId: getApiCatalog
      tags: [discovery]
      x-standard: RFC 9727
      x-observed-content-type: application/linkset+json
      x-declares:
        service-desc: https://project-hami.io/.well-known/openapi.json
        service-doc: https://project-hami.io/docs
        status: https://project-hami.io/.well-known/site-status.json
  - target: $.paths['/.well-known/agent-skills/index.json'].get
    update:
      operationId: getAgentSkillsIndex
      tags: [agents]
      x-schema: https://schemas.agentskills.io/discovery/0.2.0/schema.json
      x-skills:
        - name: hami-docs
          url: https://project-hami.io/.well-known/agent-skills/hami-docs/SKILL.md
          digest: sha256:412deb896b3f005e65c4aafeb8c2d37dbb7c188b4100f4bb3020b53905fec7a3
  - target: $.paths['/.well-known/site-status.json'].get
    update:
      operationId: getSiteStatus
      tags: [discovery]
      x-status-note: >-
        Static site metadata ({"status":"ok","service":"HAMi documentation website"}), not a live
        operational status page. Do not treat a 200 here as evidence that a HAMi deployment is healthy.
  - target: $
    update:
      x-authentication:
        scheme: none
        note: >-
          Public static documents. The site's own /.well-known/oauth-protected-resource declares an
          empty authorization_servers[] and scopes_supported[].
      x-not-described:
        - >-
          /.well-known/oauth-protected-resource is served (HTTP 200) but is not listed in this spec's
          paths — the catalog and the spec disagree by one document.
        - >-
          /sitemap.xml, /blog/rss.xml, /blog/atom.xml, /changelog/rss.xml and /changelog/atom.xml are
          all served and machine-readable but not described here.