> ## Documentation Index > Fetch the complete documentation index at: https://docs.cloudhumans.com/llms.txt > Use this file to discover all available pages before exploring further. # List proposed rewrites > The quality review queue, most impactful first. Each row pairs the answer as analysed with the rewrite proposed for it, plus the diagnosis and the performance that justify the change. ## OpenAPI ````yaml /api-reference/specs/claudia/v1.json get /v1/knowledge-bases/{kbId}/improvements/quality openapi: 3.0.1 info: title: Claudia API version: 1.0.0 servers: - url: https://api.cloudhumans.com/claudia description: Production - url: https://api.cloudhumans.com/claudia/staging description: Staging security: - bearerAuth: [] tags: - name: MCP Servers description: >- The MCP servers your agents can call tools on. Registering one here is what makes its tools selectable in an agent; the credentials it needs are stored encrypted and never read back. - name: Knowledge Base Content description: Read, search and edit the content that answers your customers. - name: Content Quality description: Rewrites proposed for content that is answering your customers badly. - name: Playground description: >- Talk to one of your agents as if you were a customer, without touching a real conversation. - name: Project Settings description: >- Read and change how a ClaudIA project behaves — the settings screens of the ClaudIA app, as an API. - name: Content Improvements description: Answers Claudia proposes for questions your content does not cover yet. - name: Knowledge Bases description: The knowledge bases your account can manage. - name: Content Sources description: The sites Claudia crawls to keep a knowledge base in sync. paths: /v1/knowledge-bases/{kbId}/improvements/quality: get: tags: - Content Quality summary: List proposed rewrites description: >- The quality review queue, most impactful first. Each row pairs the answer as analysed with the rewrite proposed for it, plus the diagnosis and the performance that justify the change. operationId: listQualityProposals parameters: - name: kbId in: path description: Knowledge base to operate on. Get the ids from listMyKnowledgeBases. required: true schema: type: string - name: status in: query description: >- Which proposals to read: the ones still waiting, the ones already decided, or all of them. required: false schema: type: string enum: - pending - approved - rejected - all - name: action in: query description: Keep only proposals that would do this to the content. required: false schema: type: string enum: - REESCREVER - MELHORAR - CONSOLIDAR - QUEBRAR - CORRIGIR_CONFIG - name: searchTerm in: query description: >- Substring match on the analysed question, the analysed answer and the proposed one. required: false schema: type: string - name: offset in: query description: Rows to skip. Page with `total` and `hasNext` from the response. required: false schema: type: integer format: int32 default: 0 example: 0 - name: limit in: query description: Page size. Capped at 200. required: false schema: type: integer format: int32 default: 50 example: 50 responses: '200': description: One page of proposals. content: application/json: schema: $ref: '#/components/schemas/QualityProposalPage' '400': description: The request could not be answered as written. content: application/json: schema: $ref: '#/components/schemas/Error' '401': description: Missing, expired or invalid credentials. '403': description: The credentials hold no account allowed to manage this content. content: application/json: schema: $ref: '#/components/schemas/Error' '404': description: >- No such resource, or none these credentials can reach — the two are deliberately indistinguishable. content: application/json: schema: $ref: '#/components/schemas/Error' '429': description: Rate limit applied by the API gateway. Back off and retry. components: schemas: QualityProposalPage: required: - hasNext - items - limit - offset - total type: object properties: items: type: array description: The proposals on this page, most impactful first. items: $ref: '#/components/schemas/QualityProposal' total: type: integer description: Total matching the query, across all pages. format: int64 example: 31 offset: type: integer description: Offset this page started at. format: int32 example: 0 limit: type: integer description: Page size actually applied (capped at 200). format: int32 example: 50 hasNext: type: boolean description: Whether another page follows. example: false description: One page of proposals. Error: required: - error type: object properties: error: type: string description: What went wrong. example: 'Forbidden: token holds no claim for the requested account' description: Something the caller needs to fix. QualityProposal: required: - action - analyzedResponse - analyzedTitle - createdAt - diagnosis - duplicateCheck - entryId - estimatedTickets - impactScore - performance - priority - reviewState type: object properties: id: type: string description: Use this id to edit, approve or reject the proposal. example: 6683f1c2a4b19e0012ab34cd entryId: type: string description: The content this proposal rewrites. Read it with getEntry. action: type: string description: >- What the proposal does to the content: REESCREVER and MELHORAR replace the answer, CORRIGIR_CONFIG also fixes the question, CONSOLIDAR and QUEBRAR restructure it and cannot be applied by this API yet. enum: - REESCREVER - MELHORAR - CONSOLIDAR - QUEBRAR - CORRIGIR_CONFIG analyzedTitle: type: string description: The question as it read when the content was analysed. analyzedResponse: type: string description: >- The answer as it read when the content was analysed. Compare it with the proposal to see what changes. proposedTitle: type: string description: Proposed new question. Only CORRIGIR_CONFIG changes it. proposedResponse: type: string description: Proposed new answer — what approving will publish. proposedChildren: type: array description: For a QUEBRAR proposal, the entries the content would be split into. items: $ref: '#/components/schemas/QualityChild' diagnosis: $ref: '#/components/schemas/QualityDiagnosis' performance: $ref: '#/components/schemas/QualityPerformance' priority: type: string description: How urgent the fix is. example: P1 recommendationSummary: type: string description: The change in one line. recommendationDetails: type: string description: The reasoning behind the change, at length. impactScore: type: number description: >- Conversations per period this content could stop losing if it answered well — the theoretical ceiling. format: double example: 42 estimatedTickets: type: number description: >- The realistic share of that ceiling, after how often the content is actually reached. format: double example: 17.5 duplicateCheck: $ref: '#/components/schemas/QualityDuplicateCheck' reviewState: type: string description: Where the proposal stands in the review. enum: - PENDING - APPROVED - REJECTED clientDecision: type: string description: The decision recorded when it was approved or rejected. enum: - APROVAR_SEM_ALTERACAO - APROVAR_COM_AJUSTES - REESCREVER_DO_ZERO - REJEITAR finalText: type: string description: >- The text actually published on approval — the proposal, or your edit of it. clientNote: type: string description: The note recorded with the decision. decidedBy: type: string description: Who decided, taken from their credentials. decidedAt: type: string description: When it was decided (ISO-8601). example: '2026-08-11T09:14:22' driftDetected: type: boolean description: >- True when the content changed after it was analysed, so the proposal was written against an older answer. False when the current answer still matches. Absent when the current answer could not be read, so drift is unknown — the proposal may be written against stale content. Review it before approving; it never blocks. example: false analysisPeriod: type: string description: The window the performance figures cover. example: 2026-03-09 → 2026-04-07 createdAt: type: string description: When the proposal was raised (ISO-8601). example: '2026-08-10T14:32:05' description: A rewrite proposed for content that is answering badly. QualityChild: required: - response - title type: object properties: title: type: string description: The question this child answers. response: type: string description: The answer. label: type: string description: Tag to apply. topic: type: string description: Topic to group it under. description: One entry a QUEBRAR proposal would split the content into. QualityDiagnosis: required: - actionability - clarity - composite - resolution type: object properties: resolution: $ref: '#/components/schemas/QualityScore' actionability: $ref: '#/components/schemas/QualityScore' clarity: $ref: '#/components/schemas/QualityScore' composite: type: number description: The three dimensions combined, 0 to 1. format: double example: 0.42 description: Why the content is being flagged. QualityPerformance: type: object properties: timesUsed: type: integer description: How many times the content was used to answer. format: int32 example: 312 retentionRatePct: type: number description: >- Share of those conversations Claudia resolved without a human, as a percentage. format: double example: 61.4 csatPct: type: number description: Customer satisfaction over those conversations, as a percentage. format: double example: 72 csatResponses: type: integer description: How many satisfaction answers that percentage is based on. format: int32 example: 38 clarifiedPct: type: number description: >- Share of conversations where the customer had to ask again to be understood, as a percentage. format: double example: 12.5 description: How the content performed over the analysed window. QualityDuplicateCheck: required: - similar - status type: object properties: status: type: string description: '`ok`, `warn` or `risk`.' example: ok note: type: string description: What the overlap is, when there is one. similar: type: array description: Existing content the rewrite may collide with. items: $ref: '#/components/schemas/SimilarEntry' description: Whether the rewrite risks overlapping existing content. QualityScore: required: - reason - score type: object properties: score: type: number description: Score from 0 (worst) to 1 (best). format: double example: 0.35 reason: type: string description: Why it scored that. description: One scored dimension, with the reasoning behind the score. SimilarEntry: required: - entryId type: object properties: entryId: type: string description: Read it with getEntry. title: type: string description: The question it answers. description: Existing content that overlaps a proposal. securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: JWT ````