> ## Documentation Index > Fetch the complete documentation index at: https://docs.momentra.org/llms.txt > Use this file to discover all available pages before exploring further. # MCP JSON-RPC 2.0 endpoint (tools/list, tools/call) > The single MCP transport endpoint, served at https://mcp.momentra.org/mcp (NOT on the REST origin). Send a JSON-RPC 2.0 request. Use `method: "tools/list"` to enumerate tools, or `method: "tools/call"` with `params.name` (one of the 24 tools) and `params.arguments` (the tool's request schema). The response is a JSON-RPC result whose `result.content[0].text` holds the tool's JSON payload (account-scoped tools also populate `result.structuredContent`). Batch-capable tools return `{ batch: true, count, items: [...] }` when given an array input. ## OpenAPI ````yaml /openapi.json post /mcp openapi: 3.1.0 info: title: Momentra Events — REST + MCP API version: 1.0.0 description: >- Momentra Events exposes local family-event data two ways in ONE document: 1. A conventional HTTP REST API under `/api/v1/*`, fronted by the ZeroClick storefront at `https://agents.momentra.org/zcj/{session}`. PATH NOTE: every REST request MUST include the full `/api/v1` prefix exactly as shown; dropping it returns 403 'Missing Authentication Token'. The `/api/v1` prefix is owned by the PATHS, not the base URL. REST auth: a ZeroClick-signed request (zc-signature) with a registered agent credential, or `Authorization: Bearer ` on read routes. REST billing: free routes do an identity check only; paid routes check allowance and settle via a `zc-usage` response header on the 200. 2. A Model Context Protocol (MCP) server at `POST /mcp` on `https://mcp.momentra.org` — a single JSON-RPC 2.0 endpoint over Streamable HTTP/SSE. It is NOT REST: every capability is an MCP `tools/call` with a `name` + `arguments`, and the 24 tools are enumerated by `tools/list`. MCP auth is OAuth 2.1 + PKCE (CIMD); some tools work anonymously (noauth). Per-tool auth + cost are noted in each `Tool_*` schema. The `/mcp` path carries a per-path `servers` override (mcp.momentra.org); the REST paths use the storefront server. The two surfaces bill the same underlying `momentra-events` service through different channels (ZeroClick per-request units vs. prepaid Momentra credits). contact: name: Momentra url: https://momentra.org servers: - url: https://agents.momentra.org/zcj/{session} description: >- REST: ZeroClick storefront (session-scoped base path) — how agents CALL the /api/v1 paths. variables: session: default: SESSION description: ZeroClick session token segment issued by the storefront. security: [] tags: - name: search description: 'REST: free, identity-scoped discovery over the global index.' - name: events description: 'REST: read event/organization data.' - name: indexing description: 'REST: index an organization and poll the job.' - name: mcp description: 'MCP: the JSON-RPC tools endpoint (tools/list, tools/call).' - name: index description: 'MCP: add organizations'' calendars and poll crawl jobs.' - name: discover description: 'MCP: search and read events/organizations already in your scope.' - name: identity description: 'MCP: connected-account profile.' - name: brief description: 'MCP: the editorial include/exclude policy (one per account).' - name: matching description: 'MCP: brief-driven discovery of events and organizations.' - name: curation description: 'MCP: accept/reject overrides that feed matching.' - name: evaluation description: 'MCP: certify a single item against the brief.' - name: seed description: 'MCP: learn a brief from a trusted aggregator org.' - name: support description: 'MCP: report a data/crawl issue to the Momentra team.' paths: /mcp: servers: - url: https://mcp.momentra.org description: >- MCP endpoint host. The /mcp JSON-RPC surface lives ONLY here, not on the REST origin. post: tags: - mcp summary: MCP JSON-RPC 2.0 endpoint (tools/list, tools/call) description: >- The single MCP transport endpoint, served at https://mcp.momentra.org/mcp (NOT on the REST origin). Send a JSON-RPC 2.0 request. Use `method: "tools/list"` to enumerate tools, or `method: "tools/call"` with `params.name` (one of the 24 tools) and `params.arguments` (the tool's request schema). The response is a JSON-RPC result whose `result.content[0].text` holds the tool's JSON payload (account-scoped tools also populate `result.structuredContent`). Batch-capable tools return `{ batch: true, count, items: [...] }` when given an array input. operationId: mcpJsonRpc requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/JsonRpcRequest' examples: toolsList: summary: List all tools value: jsonrpc: '2.0' id: 1 method: tools/list indexOrg: summary: Index an organization value: jsonrpc: '2.0' id: 1 method: tools/call params: name: index_org arguments: website: https://www.findandgoseek.net indexOrgForceHigh: summary: Index with force_high (go straight to the record/replay crawl) value: jsonrpc: '2.0' id: 1 method: tools/call params: name: index_org arguments: website: https://app.veerio.app/home effort: force_high hints: calendarUrls: - https://app.veerio.app/events eventUrls: - >- https://app.veerio.app/events/vt/middlebury/walking-tour-of-middlebury-mpxkrngv-1 updateBrief: summary: Create an active brief value: jsonrpc: '2.0' id: 1 method: tools/call params: name: update_brief arguments: brief: summary: Family-friendly events in Vermont standards: audienceTypes: - families deliveryHorizonDays: 60 geography: zipCode: '05201' radiusMiles: 50 status: active clientName: My Brief matchingEvents: summary: Find events matching the active brief value: jsonrpc: '2.0' id: 1 method: tools/call params: name: matching_events arguments: limit: 25 seedBrief: summary: Learn a brief from an aggregator org value: jsonrpc: '2.0' id: 1 method: tools/call params: name: seed_brief arguments: orgId: org_422c0a39efb3dafc responses: '200': description: >- JSON-RPC result. `result.content[0].text` is the tool payload as a JSON string; `result.structuredContent` mirrors it for account-scoped tools. On a tool-level error, `result.isError` is true. content: application/json: schema: $ref: '#/components/schemas/JsonRpcResponse' text/event-stream: schema: $ref: '#/components/schemas/JsonRpcResponse' '401': description: >- Missing/invalid OAuth token for an oauth2-only tool. The body carries an MCP www-authenticate challenge so the host can start the OAuth linking flow. security: - oauth2: - mcp - {} components: schemas: JsonRpcRequest: type: object required: - jsonrpc - method properties: jsonrpc: type: string const: '2.0' id: oneOf: - type: string - type: number method: type: string enum: - tools/list - tools/call params: type: object description: 'For tools/call: { name, arguments }.' properties: name: type: string enum: - index_org - org_status - search - fetch - search_businesses - search_business_events - list_orgs - get_profile - get_brief - update_brief - matching_events - matching_orgs - accept_event - reject_event - accept_org - reject_org - evaluate_event - evaluate_org - seed_brief - seed_status - seed_review - seed_reject - report_issue - recover_account arguments: description: The chosen tool's request object — one of the Tool_* schemas. oneOf: - $ref: '#/components/schemas/Tool_index_org' - $ref: '#/components/schemas/Tool_org_status' - $ref: '#/components/schemas/Tool_search' - $ref: '#/components/schemas/Tool_fetch' - $ref: '#/components/schemas/Tool_search_businesses' - $ref: '#/components/schemas/Tool_search_business_events' - $ref: '#/components/schemas/Tool_list_orgs' - $ref: '#/components/schemas/Tool_get_profile' - $ref: '#/components/schemas/Tool_get_brief' - $ref: '#/components/schemas/Tool_update_brief' - $ref: '#/components/schemas/Tool_matching_events' - $ref: '#/components/schemas/Tool_matching_orgs' - $ref: '#/components/schemas/Tool_curation_id' - $ref: '#/components/schemas/Tool_evaluate_id' - $ref: '#/components/schemas/Tool_seed_brief' - $ref: '#/components/schemas/Tool_seed_status' - $ref: '#/components/schemas/Tool_seed_review' - $ref: '#/components/schemas/Tool_seed_reject' - $ref: '#/components/schemas/Tool_report_issue' - $ref: '#/components/schemas/Tool_recover_account' JsonRpcResponse: type: object required: - jsonrpc properties: jsonrpc: type: string const: '2.0' id: oneOf: - type: string - type: number result: type: object description: >- MCP tool result. The tool payload is the JSON in content[0].text; structuredContent mirrors it for account-scoped tools. properties: content: type: array items: type: object properties: type: type: string const: text text: type: string description: The tool payload, JSON-encoded. structuredContent: type: object description: The tool payload as an object (account-scoped tools). isError: type: boolean error: type: object properties: code: type: integer message: type: string Tool_index_org: type: object description: >- index (oauth2; 40 credits/org fresh, 10 refresh within 30 days of a successful non-empty index, failed crawls refunded). With effort 'high'/'force_high' a successful crawl costs 80 (an extra 40 surcharge, settled only when the heavy record/replay crawl actually produced the index). Add one or more organizations' calendars via a fresh crawl. Provide `website` OR `websites` (batch up to 100). A 0-event crawl is not treated as fresh and is re-crawled next time. properties: website: type: string description: A single org website/domain. websites: type: array maxItems: 100 description: >- Batch of up to 100 entries. Each entry is EITHER a plain website/domain string OR an object { website, hints } so a caller can give different calendar/event URLs to different orgs in one call (mix freely). items: oneOf: - type: string description: A website/domain. - type: object required: - website properties: website: type: string description: The org website/domain for this entry. hints: $ref: '#/components/schemas/IndexHints' startDate: type: string description: YYYY-MM-DD, defaults to today. endDate: type: string description: YYYY-MM-DD, defaults to +90 days. effort: type: string enum: - low - high - force_high default: low description: >- Crawl effort tier. 'low' (default) runs only the fast in-stack scraper. 'high' runs the fast scraper first and escalates to a heavier record/replay crawl ONLY if the fast pass finds no events (for JS-app / API-driven sites). 'force_high' SKIPS the fast pass and goes straight to the record/replay crawl. 'high' and 'force_high' cost the same (80 on success — 40 more than low) and are slower. Applies to every org in a batch. hints: $ref: '#/components/schemas/IndexHints' description: >- Known calendar/event page URLs to AUGMENT auto-discovery (appended, never a replacement). Applies to a single `website`; for per-org hints in a batch, use the object form of `websites`. Tool_org_status: type: object required: - jobId description: >- index (noauth or oauth2; free). Poll an index job. Returns status pending|indexing|ready|failed|expired, progress+etaSeconds while running, orgId+eventCount when ready (indexStatus:'no_events_found' when a completed crawl found nothing). properties: jobId: type: string description: The jobId from index_org. Tool_search: type: object required: - query description: >- discover (noauth or oauth2; free). Search events across your indexed orgs; OpenAI search contract. Returns { results: [{ id, title, url }] } with opaque org:/event: ids for fetch. properties: query: type: string description: Free text over organizations and events. Tool_fetch: type: object description: >- discover (oauth2; 1 credit per id). Open the full document(s) for search id(s); OpenAI fetch contract. Provide `id` OR `ids` (batch up to 500). properties: id: type: string description: 'An org: or event: id from search.' ids: type: array items: type: string maxItems: 500 Tool_search_businesses: type: object required: - query description: >- discover (noauth or oauth2; free). Find your indexed organizations by name or domain. Returns { count, query, businesses: [Organization] }. properties: query: type: string description: Name or domain substring. Tool_search_business_events: type: object description: >- discover (noauth or oauth2; free). List one or more organizations' upcoming events. Provide orgId OR domain (single), or orgIds OR domains (batch up to 500). Returns { organization, eventCount, events: [Event] }. properties: orgId: type: string description: 'org: prefix tolerated.' domain: type: string orgIds: type: array items: type: string maxItems: 500 domains: type: array items: type: string maxItems: 500 Tool_list_orgs: type: object description: >- discover (noauth or oauth2; free). List every organization in your scope. Returns { count, total, organizations: [Organization] }. Unauthenticated returns an empty list. properties: limit: type: integer minimum: 1 maximum: 500 description: Max orgs to return; defaults to all in scope. Tool_get_profile: type: object description: >- identity (oauth2; free). Return the connected account's stable opaque profile { id, email?, nickname? }. No arguments. properties: {} additionalProperties: false Tool_get_brief: type: object description: >- brief (oauth2; free). Read your editorial brief. Returns { status, briefId, briefStatus, clientName?, version, updatedAt, brief } or { status:'no_brief' }. No arguments. properties: {} additionalProperties: false Tool_update_brief: type: object description: >- brief (oauth2; free; write, not destructive). Create/update your brief. `brief` REPLACES the stored policy document; `status` active is required for matching to use it. properties: brief: $ref: '#/components/schemas/Brief' status: type: string enum: - draft - active description: >- draft = not used for matching; active = used by matching_events/matching_orgs. clientName: type: string description: A label for the brief. Tool_matching_events: type: object description: >- matching (oauth2; free). Events that match your ACTIVE brief (found FOR you): brief embedded as the query, ranked by similarity, facet-filtered, freshness-windowed, curation-honored, scoped to entitled orgs. Returns { status, briefId, count, matches:[{id,score,title,url,orgId,startEpoch,facets}] }; no_active_brief when none. properties: limit: type: integer minimum: 1 maximum: 100 description: Max matches (default 25). Tool_matching_orgs: type: object description: >- matching (oauth2; free). Organizations that match your active brief. Returns { status, briefId, count, matches:[{id,orgId,score,name,domain,categories?,inclusivity?}] }; requires an active brief. properties: limit: type: integer minimum: 1 maximum: 100 description: Max matches (default 25). Tool_curation_id: type: object required: - id description: >- curation (oauth2; free; write, not destructive). accept_event / reject_event / accept_org / reject_org. accept pins an item into matching; reject drops it. `id` is a single id or an array (batch). Event ids are event::; org ids are org:. Returns { status:'ok', decision, kind, count, ids }. properties: id: oneOf: - type: string - type: array items: type: string minItems: 1 Tool_evaluate_id: type: object required: - id description: >- evaluation (oauth2; free; read-only). evaluate_event / evaluate_org. Certify item(s) against the active brief: exclude rules + facet gate, then similarity. `id` is a single id or an array. Returns per item { status, id, verdict (fit|weak|reject), similarity?, rulesPass, facetPass, reasons[] }; requires an active brief. properties: id: oneOf: - type: string - type: array items: type: string minItems: 1 Tool_seed_brief: type: object description: >- seed (oauth2; free; write, not destructive). Learn a brief from a trusted aggregator org (events mostly link off-domain, >25% required). Provide orgId OR website. Returns { status:'seeding', jobId, orgId, offDomainPct, totalEvents } or { status:'not_aggregator', ... }. On success a background job converges a brief and sets it active — poll seed_status. properties: orgId: type: string description: An indexed aggregator org id. website: type: string description: Aggregator website (resolved to an orgId). Tool_seed_status: type: object required: - jobId description: >- seed (oauth2; free; read-only). Poll a seed_brief job. Returns { status (seeding|reviewing|ready|failed|not_aggregator), jobId, orgId, iteration, maxIterations, fitPct, residualCount, totalEvents, briefId?, message }. properties: jobId: type: string Tool_seed_review: type: object required: - jobId description: >- seed (oauth2; free; read-only). Pull borderline events to review during/after seeding (near the brief but not clearly on it), re-ranked against the latest brief, decided items excluded. Returns { status, jobId, seedState, count, candidates:[{id,title,orgId,domain,score,reason}] }. properties: jobId: type: string limit: type: integer minimum: 1 maximum: 50 description: Max candidates (default 25). Tool_seed_reject: type: object required: - jobId - ids description: >- seed (oauth2; free; write, not destructive). Submit events rejected as off-brief during a seed review. Records durable negatives AND returns a proposed brief tightening + reasoning + clarifying questions (guarded so it never blocks the aggregator's own events). Apply via update_brief. Returns { status:'ok', jobId, rejectedCount, curationWritten, proposedPatch, rationale[], questions[], droppedForGuard[] }. properties: jobId: type: string ids: type: array items: type: string minItems: 1 description: Event ids (event::) to reject. Tool_report_issue: type: object required: - message description: >- support (noauth or oauth2; free; write, not destructive). Report a problem with Momentra data or a crawl (missing events, wrong venue/address, an org failing to parse). Emails the Momentra team with the message plus auto-collected diagnostics (resolved org, indexed eventCount + address, scraper scriptVersion, sample events, related job status). Works anonymously; reporter identity is attached when the connection is linked. Returns { status:'received', reportId, delivered, message }. properties: message: type: string description: What you're seeing (required). domain: type: string description: The org website/domain the issue is about, if any. orgId: type: string description: The Momentra orgId the issue is about, if known. jobId: type: string description: >- A related index/scrape jobId, if the issue concerns a specific crawl. email: type: string description: Where to reply. Optional if the connection is linked. Tool_recover_account: type: object description: >- identity (oauth2; free; write, not destructive). Recover a previous account's purchased credits after reconnecting with a new login. Takes NO arguments — returns a LINK (recoveryUrl) to the Momentra /recover/start page. On that page the user enters the email tied to the prior account; a confirmation LINK is then emailed to that address (no email or code typed into chat — works in connectors that can't relay either). Opening the emailed link on the /recover page and confirming REBINDS the current connection onto the recovered account (its balance + history kept) and consolidates duplicate same-email accounts into it. Same-email gated; the emailed link is single-use, 15-min TTL, rate-limited; nothing is duplicated and the free-signup grant is never re-run. Returns { status: start | error, message, recoveryUrl }. properties: {} IndexHints: type: object description: >- Caller-supplied discovery hints appended to (never replacing) auto-discovered URLs. Both lists are http(s)-only, deduped, and capped at 50. properties: calendarUrls: type: array items: type: string maxItems: 50 description: Known calendar/listing page URLs. eventUrls: type: array items: type: string maxItems: 50 description: Known single-event detail page URLs. Brief: type: object description: >- The editorial policy document. All fields optional; an empty facet group means 'no restriction'. update_brief replaces it wholesale. properties: summary: type: string description: >- Semantic query text: matching embeds this + eventKinds.include and ranks by similarity. standards: type: object description: >- Closed-vocabulary facet filters (enabled values allowed through). Fail-open: events missing a facet are never dropped. properties: ageBands: type: array items: type: string enum: - general - babies - toddlers - kinder - kids - tweens - teens - all_ages audienceTypes: type: array items: type: string enum: - families - children_with_adult - adults - seniors - students - professionals commercialIntent: type: array items: type: string enum: - programmed_event - promotion_only - unknown religiousContent: type: array items: type: string enum: - worship_or_service - faith_related_non_service - none - unknown accessRestrictions: type: array items: type: string enum: - registration_required - ticket_required - members_only - students_only - age_restricted - invite_only eventCategories: type: array items: type: string enum: - live_music - food_drink - performing_arts - outdoors - family_friendly - educational - fundraiser programmingNature: type: array items: type: string enum: - performance - class_or_workshop - social_gathering - meeting - exhibit - market_or_sale - fundraiser - other deliveryHorizonDays: type: integer description: Freshness window in days. sourceDeny: type: array items: type: string description: Source domains to exclude. sourceAllow: type: array items: type: string eventKinds: type: object properties: include: type: array items: type: string exclude: type: array items: type: string geography: type: object description: Hard geo gate. properties: zipCode: type: string radiusMiles: type: number excludePhrases: type: array items: type: string description: Deterministic title-phrase excludes. securitySchemes: oauth2: type: oauth2 description: >- MCP auth: OAuth 2.1 + PKCE with Client ID Metadata Documents (CIMD). Authorization-server metadata (RFC 8414) and protected-resource metadata (RFC 9728) are served from the MCP endpoint origin (mcp.momentra.org). flows: authorizationCode: authorizationUrl: https://mcp.momentra.org/authorize tokenUrl: https://mcp.momentra.org/token scopes: mcp: Access the Momentra Events MCP tools for the linked account. ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.