generated: '2026-09-19'
method: derived
source: mcp/getaiscan-app-mcp-descriptor.json + a2a/getaiscan-app-agent-card.json + openapi/getaiscan-app-openapi.json
summary: >-
  AIScan has one execution surface (the x402 REST API, 20 operations) and three agent-facing declarations over
  it: an MCP discovery descriptor declaring one tool, an A2A agent card declaring 18 skills, and an ai-plugin
  manifest pointing at the OpenAPI. This crosswalk binds every declared tool and skill to the OpenAPI
  operationId that fulfils it. Confidence is high throughout because the bindings are explicit in the source
  documents: each A2A skill's x-payment-info.endpoint IS the REST route, and the skill ids equal the
  operationIds verbatim. The one MCP tool binds to full_audit by live observation (POST /api/agent/scan
  returned the full_audit 402 challenge). There is no live MCP tools/list to reconcile — no MCP server exists
  (see mcp/getaiscan-app-mcp.yml).
surfaces:
  openapi:
    file: openapi/getaiscan-app-openapi.json
    source: https://api.getaiscan.app/openapi.json
    operations: 20
    gated: false — the document is public; every paid operation answers HTTP 402 x402 until paid
  mcp:
    descriptor: https://getaiscan.app/.well-known/mcp.json
    server: null
    tools_list: not available — no MCP endpoint exists; the descriptor's static tools[] is the only tool list
    gated: n/a
  a2a:
    card: https://getaiscan.app/.well-known/agent-card.json
    endpoint: https://api.getaiscan.app/a2a
    skills: 18
    note: message/send returns routing guidance and does not execute skills; skills execute over REST.
  graphql: null
crosswalk:
- tool: scan_website
  surface: mcp-descriptor
  category: audit
  rest: [full_audit]
  binding: rest
  confidence: high
  note: >-
    The descriptor points the tool at https://api.getaiscan.app/api/agent/scan; a live POST to that route
    returned the x402 challenge for resource /api/agent/full_audit (0.35 USDC). The tool's inputSchema {url}
    equals full_audit's requestBody schema; its outputSchema (overall score/grade, four scores, categories) is
    richer than the generic response schema the OpenAPI declares for full_audit.
- {tool: check_health, surface: a2a-skill, category: check, rest: [check_health], binding: rest, confidence: high}
- {tool: check_llms_txt, surface: a2a-skill, category: check, rest: [check_llms_txt], binding: rest, confidence: high}
- {tool: check_schema, surface: a2a-skill, category: check, rest: [check_schema], binding: rest, confidence: high}
- {tool: check_agent_files, surface: a2a-skill, category: check, rest: [check_agent_files], binding: rest, confidence: high}
- {tool: check_mcp, surface: a2a-skill, category: check, rest: [check_mcp], binding: rest, confidence: high}
- {tool: score_aeo, surface: a2a-skill, category: score, rest: [score_aeo], binding: rest, confidence: high}
- {tool: score_geo, surface: a2a-skill, category: score, rest: [score_geo], binding: rest, confidence: high}
- {tool: score_agent, surface: a2a-skill, category: score, rest: [score_agent], binding: rest, confidence: high}
- {tool: score_mcp, surface: a2a-skill, category: score, rest: [score_mcp], binding: rest, confidence: high}
- {tool: full_audit, surface: a2a-skill, category: audit, rest: [full_audit], binding: rest, confidence: high}
- {tool: generate_llms_txt, surface: a2a-skill, category: generate, rest: [generate_llms_txt], binding: rest, confidence: high}
- {tool: generate_mcp_json, surface: a2a-skill, category: generate, rest: [generate_mcp_json], binding: rest, confidence: high}
- {tool: fix_pack, surface: a2a-skill, category: fix, rest: [fix_pack], binding: rest, confidence: high}
- {tool: compare, surface: a2a-skill, category: compare, rest: [compare], binding: rest, confidence: high, note: 'body needs url + competitor'}
- {tool: full_report, surface: a2a-skill, category: bundle, rest: [full_report], binding: rest, confidence: high}
- {tool: visibility_check, surface: a2a-skill, category: brand-visibility, rest: [visibility_check], binding: rest, confidence: high, note: 'body needs brand + niche; no url'}
- {tool: visibility_vs_competitor, surface: a2a-skill, category: brand-visibility, rest: [visibility_vs_competitor], binding: rest, confidence: high, note: 'body needs brand + competitor_brand + niche'}
- {tool: visibility_fix_pack, surface: a2a-skill, category: brand-visibility, rest: [visibility_fix_pack], binding: rest, confidence: high, note: 'body needs brand + niche + weak_spots[] from a prior visibility_check'}
mcp_only: []
a2a_only: []
rest_only:
- operationId: generate_skill
  method: POST
  path: /api/agent/generate_skill
  price_usdc: '0.45'
  reason: Added in OpenAPI 5.9.4 and present in the live index; the agent card (5.9.1) and the descriptor predate it.
- operationId: index
  method: GET
  path: /api/agent/index
  price_usdc: free
  reason: The free catalog; a discovery call rather than a capability, so no skill or tool wraps it.
coverage:
  mcp_tools_declared: 1
  mcp_tools_bound: 1
  a2a_skills_declared: 18
  a2a_skills_bound: 18
  mcp_only: 0
  a2a_only: 0
  rest_operations_total: 20
  rest_operations_with_a_tool_or_skill: 18
  rest_only: 2