> ## Documentation Index > Fetch the complete documentation index at: https://docs.openpulse.cloud/llms.txt > Use this file to discover all available pages before exploring further. # Get one signal > A single signal by id, regardless of how old it is or where it falls in the list's pagination. Use this to resolve a signal a link names; use `GET /v1/results` to browse. ## OpenAPI ````yaml /api-reference/openapi.json get /v1/results/{id} openapi: 3.1.0 info: title: Open Pulse API version: 1.0.0 summary: Read buying signals, accounts and pipeline from an Open Pulse workspace. description: >- The programmatic surface of Open Pulse, authenticated with a workspace API key. **A key always acts as a member, never an admin.** Privilege comes from scopes, and the scope vocabulary has no way to express an admin action: a leaked key can read signals and move deals, and can never create a listener, change billing, or mint another key. Routes that declare no scope are unreachable by a key at all, which is why this document describes fewer operations than the server has handlers. For an assistant rather than a program, prefer the MCP server at `/mcp`: it is the same data with tool descriptions a model can act on, and it uses OAuth rather than a shared secret. See `/api.md`. **Webhooks**, described below, are the other direction: calls this workspace's server makes to yours. Verify each one by recomputing the HMAC and comparing it to the `x-openpulse-signature` header — an unverified delivery is a POST from anyone who finds the URL, not proof it came from Open Pulse. contact: email: support@openpulse.cloud url: https://openpulse.cloud/api.md license: name: Proprietary identifier: LicenseRef-OpenPulse-Terms servers: - url: https://api.openpulse.cloud description: Production security: - apiKey: [] externalDocs: url: https://openpulse.cloud/api.md description: Open Pulse for agents paths: /v1/results/{id}: get: tags: - results summary: Get one signal description: >- A single signal by id, regardless of how old it is or where it falls in the list's pagination. Use this to resolve a signal a link names; use `GET /v1/results` to browse. operationId: getV1ResultsById parameters: - name: id in: path required: true description: The signal's id. schema: type: string responses: '200': description: >- The signal, outside every window and every page. Unlike the list, this applies no `postedAt` bound — a lead found yesterday but published five weeks ago is returned here and is not in the list that links to it. content: application/json: schema: $ref: '#/components/schemas/Signal' '401': description: The key is missing, malformed, expired or revoked. content: application/json: schema: $ref: '#/components/schemas/Error' '403': description: The key lacks the `signals:read` scope. content: application/json: schema: $ref: '#/components/schemas/Error' '404': description: >- No signal with this id, or it belongs to another workspace. The two are deliberately indistinguishable. content: application/json: schema: $ref: '#/components/schemas/Error' '429': description: >- Too many requests from this key. `Retry-After` carries the seconds to wait. content: application/json: schema: $ref: '#/components/schemas/Error' security: - apiKey: - signals:read components: schemas: Signal: type: object description: >- One public post, job posting or review a listener judged relevant, with its score and the reasoning behind it. **Most fields are optional and that is structural, not incidental.** Rows are stored by a pipeline that has gained stages over time, and a field added after a row was written is absent from it for ever: nothing is re-scored, because that is a model call per signal for no new information. Read defensively. required: - id - orgId - listenerId - source - url - title - excerpt - postedAt - relevanceScore - classification - rationale - matchedKeywords - firstSeenAt properties: id: type: string description: Stable identifier. What `GET /v1/results/{id}` takes. orgId: type: string description: The workspace this belongs to (`org_…`). listenerId: type: string description: The listener that found it. source: type: string enum: - x - linkedin - facebook - reddit - hackernews - linkedin_jobs - indeed - ashby - greenhouse - lever - google_reviews - g2 - trustpilot - youtube description: The platform it came from. url: type: string description: Link to the post itself. format: uri title: type: string description: The post's title, or its first line where the platform has none. excerpt: type: string description: >- The post's text, as much of it as the source gave up. See `coverage`. author: type: string description: The poster's handle, where the source reports one. publishedAt: type: string format: date-time description: When the post was published, when the source reports it. postedAt: type: string format: date-time description: >- The sort and filter key: `publishedAt` when known, otherwise `firstSeenAt`. approximateDate: type: boolean description: >- True when `publishedAt` was derived from "3 months ago" rather than a real timestamp. Rendered as `~`, because implying a precision the source never gave is worse than admitting the approximation. dedupeKey: type: string description: >- Normalised URL, used to collapse one post reached through different links. titleKey: type: string description: Normalised title, used to collapse one post syndicated across URLs. coverage: type: string enum: - full_text - snippet - best_effort - enriched description: >- How much of the post the classifier actually read. `snippet` and `best_effort` cap the evidence component of the score; `enriched` means the page behind a snippet was fetched. Absent on rows written before the enrichment stage. relevanceScore: type: integer description: 0–100. The sort order of every list in the product. minimum: 0 maximum: 100 scoreVersion: type: integer enum: - 1 - 2 - 3 description: >- Which rubric produced `relevanceScore`. No version is comparable to another and they sort against each other in the same column, which is why they are marked rather than silently mixed. Absent means 1. scoreBreakdown: $ref: '#/components/schemas/ScoreBreakdown' classification: type: string enum: - high_intent - evaluating - relevant - competitor - content - strong_match - worth_a_look - stretch - mismatch - stale - hiring_now - expanding - replacing - adjacent - switching - frustrated - comparing - praise - pain_confirmed - at_risk - healthy - noise description: >- The intent label. **The valid set depends on the listener's mode** — a `market` listener never produces `pain_confirmed` — so filtering by one from the wrong mode returns an empty page rather than an error. The retired `jobs` labels are still valid here: no listener produces them, and signals carrying them are still stored and still read. signalCategory: type: string enum: - direct_intent - hire_vs_buy - competitor_pain - growth - funding - procurement - construction - regulatory - operational_pain - tech_change - leadership_change - real_estate - new_business - cx_failure - missing_capability - incident description: >- What kind of event this is, as opposed to how good it is. Orthogonal to `classification`: a funding round and a support complaint can both be `high_intent`, and are not actioned by the same person. urgency: type: string enum: - low - medium - high - immediate description: How soon this needs acting on. purchaseIntent: type: string enum: - weak - moderate - strong - explicit description: How directly a wish to buy was expressed. confidence: type: string enum: - low - medium - high description: How much the model trusts its own read. inferredNeed: type: string description: 'What the subject probably needs: the product-shaped restatement.' recommendedAction: type: string description: The single next action worth taking. situation: type: string description: >- What is happening, stated without reference to the seller. Separate from `opportunity` so the leap between them is checkable by whoever reads the row. opportunity: type: string description: >- Why that situation matters to this seller, or absent when it does not. inferenceHops: type: integer enum: - 0 - 1 description: >- How far the reasoning reached: `0` they said it, `1` one step from what they said. There is no `2` — two stacked assumptions are indistinguishable from invention, and the prompt refuses it. rationale: type: string description: Why this was kept, in the model's words. suggestedAngle: type: string description: 'One line on why this person might care: the outreach angle.' matchedKeywords: type: array items: type: string description: A keyword from the listener's plan. description: Computed from the text, never taken from the model. themes: type: array items: type: string description: A short noun phrase. description: Two to four phrases naming what the post is about. rating: type: number description: >- Star rating 1–5 on a review source. Never the same number as `relevanceScore`. minimum: 1 maximum: 5 subjectOrg: $ref: '#/components/schemas/SubjectOrg' subjectName: type: string description: Display name of the place or rival, so a row reads without a join. accountKey: type: string description: >- The company this is about, namespaced (`d:acme.com` for a domain, `n:` for a name, `p:` for a place). **Absent is a real and common answer** — a thread that names no company has none. rivalKey: type: string description: Which rival this is about, on a `rivals` listener. placeKey: type: string description: Which place this review belongs to, on a `places` listener. policy: $ref: '#/components/schemas/SurfacingPolicy' passedBy: type: string enum: - keyword - rerank description: >- How this got past the on-topic gate. `rerank` means it shares no keyword and the cross-encoder judged it relevant anyway. rerankScore: type: number description: Cross-encoder relevance, when the rerank stage ran. foundByQuery: type: string description: Which planned query surfaced this, for per-query yield reporting. queryAttribution: type: string enum: - exact - inferred description: >- Whether `foundByQuery` is known or was worked out afterwards from token overlap. Recorded because per-query yield decides which queries get retired. assignedTo: type: object description: >- Who is working this lead. Set through `PATCH /v1/results/{id}/assign`. required: - userId - at - by properties: userId: type: string description: Clerk user id of the assignee. at: type: string format: date-time description: When it was assigned. by: type: string description: Clerk user id of whoever assigned it. feedback: type: object description: A person's verdict on the signal, which the classifier reads back. required: - verdict - at properties: verdict: type: string enum: - good - bad description: What they said. at: type: string format: date-time description: When they said it. readAt: type: string format: date-time description: >- When somebody in the workspace read it. **Workspace-wide rather than per member**: a signal is a lead somebody acts on once. Absent means unread. alertedAt: type: string format: date-time description: When an urgency alert was sent for this. Absent means never. firstSeenAt: type: string format: date-time description: >- When this workspace first saw it, which is not when it was published. Error: type: object description: What went wrong, in a shape every failing route shares. required: - message properties: message: type: string description: Human-readable explanation, safe to show a user. code: type: string description: >- Stable machine-readable error code, when the failure has one worth branching on. `website_unreadable` and `forbidden` are the two a caller acts on. detail: description: >- Validator output on a 400, as an array of issues naming the offending field. Diagnostic only: the shape comes from the validation library and is not part of the contract. ScoreBreakdown: type: object description: >- The seven components the score is the sum of, each capped at its own weight. Present only on `scoreVersion` 2 and 3. The components are the point rather than the total: "86" says nothing, "26 of 30 on intent, 8 of 20 on fit" says a real buyer who is probably not yours. required: - purchaseIntent - fit - urgency - recency - commercialValue - evidenceQuality - multiSignal properties: purchaseIntent: type: integer description: 0–30. minimum: 0 maximum: 30 fit: type: integer description: 0–20. minimum: 0 maximum: 20 urgency: type: integer description: 0–15. minimum: 0 maximum: 15 recency: type: integer description: 0–10. minimum: 0 maximum: 10 commercialValue: type: integer description: 0–10. minimum: 0 maximum: 10 evidenceQuality: type: integer description: 0–10. minimum: 0 maximum: 10 multiSignal: type: integer description: 0–5. minimum: 0 maximum: 5 SubjectOrg: type: object description: >- The organisation the signal is about, which is not the poster. Every field is optional and absent is the common case: most conversations name no company, and inventing one from the platform host would file the whole corpus under one domain. properties: name: type: string description: The company named in the post. location: type: string description: Free text as written, which `region` on the account is parsed from. workMode: type: string enum: - any - remote - hybrid - onsite description: Where the work happens, on a posting. seniority: type: string description: Only meaningful on a posting. compensation: type: string description: As stated, when stated. Never normalised. SurfacingPolicy: type: object description: >- What the surfacing policy did when this signal was stored, for off-policy evaluation. Absent on every signal written before the holdout existed — and absent must read as "unknown", never as a propensity of 1. required: - propensity - explored - scoreAtDecision - barAtDecision properties: propensity: type: number description: P(shown) under the policy in force at the time. Never 0. exclusiveMinimum: 0 maximum: 1 explored: type: boolean description: True when the random holdout surfaced this rather than its score. scoreAtDecision: type: integer description: The score this was judged on. barAtDecision: type: integer description: The listener's bar at that moment. securitySchemes: apiKey: type: http scheme: bearer description: >- A workspace API key, created in workspace settings and sent as `Authorization: Bearer op_live_…`. Session cookies are not accepted. Scopes: signals:read, accounts:read, listeners:read, listeners:analyse, pipeline:read, pipeline:write, collections:write, export:read. ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.