generated: '2026-09-19'
method: derived
source: >-
  mcp/optionsahoy-com-mcp-tools.json (live tools/list, 2026-09-19) and a2a/optionsahoy-com-agent-card.json
  (live card) bound to openapi/optionsahoy-com-openapi.json (OpenAPI 3.1.0, 11 operations, fetched from
  https://optionsahoy.com/openapi.json). The provider states the binding itself: llms-full.txt prefixes every
  tool with "REST equivalent: POST https://optionsahoy.com/api/v1/<path>", and the for-agents page says
  "Every endpoint under /api/v1/* is wired to the same engine an MCP tool call lands on, so the response your
  agent would receive is byte-identical".
purpose: >-
  Make the MCP tool the first-class discovery unit and bind it to its backing REST operation. OptionsAhoy is
  the rare case where the three agent surfaces are the SAME eight calculators with the same names and the same
  input schemas; the divergence is in what surrounds them (MCP adds resources, prompts and a widget; REST adds
  discovery, stats and a badge; A2A adds nothing beyond the eight skills).
surfaces:
  openapi:
    file: openapi/optionsahoy-com-openapi.json
    url: https://optionsahoy.com/openapi.json
    operations: 11
    gated: false
    note: Every operation carries an operationId, a summary, a description and a tag. Eight are the calculators; three are discovery/telemetry.
  mcp:
  - url: https://optionsahoy.com/mcp
    gated: false
    note: Anonymous tools/list returned 8 tools with inputSchema, outputSchema and annotations on 2026-09-19.
  graphql: null
  a2a:
    url: https://optionsahoy.com/a2a
    card: a2a/optionsahoy-com-agent-card.json
    gated: false
    protocol_version: '0.3.0'
    skills: 8
    note: Skill ids are exactly the MCP tool names; a skill is run with a DataPart {"skill":"<id>","input":{...}} whose input is the same JSON body the REST operation takes.
crosswalk:
- tool: amt_iso_optimize
  surface: [mcp, a2a]
  category: iso-amt
  rest: [optimizeAmtIso]
  rest_path: POST /api/v1/amt-iso
  binding: rest
  confidence: high
  input_schema: components.schemas.AmtIsoInput (9 required of 16 properties)
  output_schema: components.schemas.AmtIsoResult
- tool: nso_calculate
  surface: [mcp, a2a]
  category: nso
  rest: [calculateNso]
  rest_path: POST /api/v1/nso
  binding: rest
  confidence: high
  input_schema: components.schemas.NsoInput (9 required of 14)
  output_schema: components.schemas.NsoResult
- tool: rsu_sell_vs_hold
  surface: [mcp, a2a]
  category: rsu
  rest: [calculateRsu]
  rest_path: POST /api/v1/rsu-sell-vs-hold
  binding: rest
  confidence: high
  input_schema: components.schemas.RsuInput (7 required of 12)
  output_schema: components.schemas.RsuResult
- tool: concentration_analyze
  surface: [mcp, a2a]
  category: concentration
  rest: [calculateConcentration]
  rest_path: POST /api/v1/concentration
  binding: rest
  confidence: high
  input_schema: components.schemas.ConcentrationInput (8 required of 14)
  output_schema: components.schemas.ConcentrationResult
- tool: protective_put_price
  surface: [mcp, a2a]
  category: hedging
  rest: [priceProtectivePut]
  rest_path: POST /api/v1/protective-put
  binding: rest
  confidence: high
  input_schema: components.schemas.ProtectivePutInput (4 required of 9)
  output_schema: components.schemas.ProtectivePutResult
- tool: qsbs_check
  surface: [mcp, a2a]
  category: qsbs
  rest: [checkQsbs]
  rest_path: POST /api/v1/qsbs
  binding: rest
  confidence: high
  input_schema: components.schemas.QsbsInput (12 required of 12)
  output_schema: components.schemas.QsbsResult
- tool: equity_funding_plan
  surface: [mcp, a2a]
  category: equity-funding
  rest: [planEquityFunding]
  rest_path: POST /api/v1/equity-funding
  binding: rest
  confidence: high
  input_schema: components.schemas.EquityFundingInput (5 required of 12)
  output_schema: components.schemas.EquityFundingResult
- tool: rsu_lot_optimize
  surface: [mcp, a2a]
  category: rsu-lot-order
  rest: [optimizeRsuLotOrder]
  rest_path: POST /api/v1/rsu-lot-order
  binding: rest
  confidence: high
  input_schema: components.schemas.RsuLotOptimizeInput (7 required of 7)
  output_schema: components.schemas.RsuLotOptimizeResult
  note: The REST operation is tagged RsuLotOptimize, a tag the spec's tags[] does not declare (see overlays/optionsahoy-com-openapi-overlay.yaml).
mcp_only: []
mcp_only_surfaces:
- {kind: resources, count: 9, reason: 'Eight editorial/learn pages served as text/markdown plus the ui://widget/optionsahoy-scenario.html Apps-SDK widget; no REST operation returns them.'}
- {kind: prompts, count: 8, reason: 'One prompt per calculator plus plan-equity-portfolio, a composite that sequences several tools; REST has no prompt surface.'}
- {kind: session, reason: 'Mcp-Session-Id assigned at initialize and tools/call _meta.optionsahoy next-steps; the REST twin is the next_steps field in each response body.'}
rest_only:
- {operationId: discover, path: GET /api/v1, capability: discovery, note: 'Endpoint inventory; the MCP twin is tools/list and the GET descriptor on /mcp.'}
- {operationId: stats, path: GET /api/v1/stats, capability: telemetry, note: 'Public aggregate MCP usage counts (211,760 total calls observed 2026-09-19).'}
- {operationId: badge, path: GET /api/v1/badge, capability: telemetry, note: 'shields.io endpoint-badge JSON built from the same stats.'}
a2a_skills:
  note: All eight A2A skills bind 1:1 to the rows above by id; the A2A surface carries no skill that is not also an MCP tool and a REST operation.
coverage:
  tools_named: 8
  tools_bound: 8
  mcp_only: 0
  rest_operations_total: 11
  rest_operations_with_tool: 8
  rest_only: 3
  a2a_skills: 8
  a2a_skills_bound: 8