> ## Documentation Index > Fetch the complete documentation index at: https://docs.erdo.ai/llms.txt > Use this file to discover all available pages before exploring further. # DeployPageAPI mirrors the erdo_deploy_page MCP tool. clone_from_page_id in > the body switches it into clone mode — a new page copied server-side from an existing one, instead of content supplied here. ## OpenAPI ````yaml /api/openapi.json post /v1/pages openapi: 3.0.0 info: title: Erdo API description: >- Erdo's REST API: query and write datasets, run agents, manage threads, integrations, pages, evals, workstreams, experiments, and bounded outreach. Authenticate with a Bearer API key (erdo_api_...) or scoped token (erdo_token_...). version: '2026-09-09' servers: - url: https://api.erdo.ai description: Production security: - bearerAuth: [] paths: /v1/pages: post: summary: > DeployPageAPI mirrors the erdo_deploy_page MCP tool. clone_from_page_id in description: > the body switches it into clone mode — a new page copied server-side from an existing one, instead of content supplied here. operationId: POST:mcp.DeployPageAPI requestBody: content: application/json: schema: properties: clone_from_page_id: description: > byte-identical server-side copy of the named page rather than content the caller supplies. It is a parameter here, not a separate endpoint/tool, for the same reason a WHERE-clause variant is a parameter: cloning is deploying a page whose content comes from another page. title: > CloneFromPageID switches the deploy into CLONE mode: the new page is a type: string css: type: string dataset_slugs: items: type: string type: array html: type: string js: type: string kv_slugs: items: type: string type: array meta_description: type: string meta_icon_asset_id: description: > knowledge attachment ids (from the org's Brand Style record) the public shell renders as the browser-tab icon and the share-card image. The page itself is an iframe, so this is the only way its tab/preview can carry the customer's brand. Optional; unset falls back to Erdo's icon and card. title: > MetaIconAssetID / MetaImageAssetID are the page's PUBLIC branding: the type: string meta_image_asset_id: type: string meta_title: description: > title + description shown when the page is shared or crawled) — distinct from Title, the internal library name. Mirrors the create\_html\_artifact convention in backend/agent/agents/helpers.go. title: > MetaTitle / MetaDescription are the page's PUBLIC SEO metadata (OG/meta type: string public: type: boolean runtime: type: string title: description: > required — in clone mode both must be absent. Content mode still requires them; deployPageLogic enforces it with a clear error. title: > Title and HTML carry json omitempty so the MCP schema does not mark them type: string writable_dataset_slugs: description: > named-KV-store access that erdo.insertRows and erdo.kv.set need. Without them those writes 403 ("this page has no write access"). dataset\_slugs is read-only; these are the write side. items: type: string title: > WritableDatasetSlugs / KVSlugs / WritableKVSlugs grant the write and type: array writable_kv_slugs: items: type: string type: array type: object responses: '200': content: application/json: schema: properties: cloned_pipelines: items: $ref: '#/components/schemas/mcp.ClonedPipelineSummary' type: array id: type: string meta_description: type: string meta_icon_asset_id: description: | ids (favicon / share image), nil when unset. title: > MetaIconAssetID / MetaImageAssetID echo the page's branding attachment type: string meta_image_asset_id: type: string meta_title: description: > (nil when unset), so a caller can confirm what it just deployed/updated without a separate read. title: > MetaTitle / MetaDescription echo the page's current public SEO metadata type: string notes: items: type: string type: array public: type: boolean public_url: type: string review: $ref: '#/components/schemas/agents.AgentArtifactReview' rewrites: additionalProperties: type: string type: object source_page_id: description: > clone\_from\_page\_id deploy: the page the copy was made from, the lead-capture pipelines duplicated onto it, every id substitution applied to the copied content, and anything the caller still has to act on. title: > SourcePageID / ClonedPipelines / Rewrites / Notes are set only on a type: string thread_id: type: string title: type: string url: description: > share link and is only present while the page is public. Use these values verbatim — hosts differ across prod/staging/local. title: > URL is the authenticated editor view; PublicURL is the visitor-facing type: string validation: $ref: '#/components/schemas/types.ArtifactValidationResult' warning: description: > successful content deploy/update). The page exists and is usable; the calling agent should surface this to the user and retry the failing op. title: > Warning carries non-fatal issues (e.g. visibility update failed after a type: string type: object description: Success response default: $ref: '#/components/responses/APIError' components: schemas: mcp.ClonedPipelineSummary: description: > the copy: source\_id is the pipeline the original page posts to, id the copy's own pipeline. purpose/slug identify it the way the rest of the pipeline surface does. properties: id: type: string purpose: type: string slug: type: string source_id: type: string title: > ClonedPipelineSummary reports one lead-capture event pipeline duplicated onto type: object agents.AgentArtifactReview: description: > request\_review. Structured rather than prose so the tool result can carry it without a second formatting layer inventing its own vocabulary. properties: issues: items: $ref: '#/components/schemas/agents.ArtifactCriticIssue' type: array judges: items: type: string type: array lighthouse: $ref: '#/components/schemas/types.PageReviewLighthouse' reason: type: string reviewed: type: boolean summary: type: string title: > AgentArtifactReview is what an agent gets back when it asks for a review with type: object types.ArtifactValidationResult: description: > action callers can reference it without importing the validator implementation. properties: errors: items: $ref: '#/components/schemas/types.ArtifactValidationError' type: array valid: type: boolean title: > ArtifactValidationResult lives in this package so both the private API layer and type: object agents.ArtifactCriticIssue: properties: category: title: > brand | technical | conversion | copy | responsive | accessibility | compliance type: string description: title: | what's wrong, quoting the offending markup type: string id: title: | e.g. "brand-1", "technical-2" type: string judge_slug: description: > flat merged result fans back into per-judge verdicts (SplitCriticResultByJudge). Empty on the single-rubric path, where the call itself IS one judge. title: > JudgeSlug tags a MERGED-mode issue with the rubric it was raised against, so a type: string severity: title: | low | medium | high | critical type: string suggestion: title: | the concrete change to make type: string title: | ArtifactCriticIssue is one fixable problem found in the page. type: object types.PageReviewLighthouse: description: > the PageSpeed Insights API on the mobile strategy. Each is 0..100 (Lighthouse's own 0..1 category score ×100, rounded). Optional by design: absent when PSI is unavailable, rate-limited, or times out — a missing Lighthouse never fails the review. FetchedAt stamps when the measurement was taken. Each category is a POINTER because Lighthouse itself reports null for a category whose lab run failed (observed live: performance=null on a page the other three categories scored fine) — a non-pointer int rendered that null as a fake catastrophic 0. An absent category is absent, never zero. properties: accessibility: format: int64 type: integer best_practices: format: int64 type: integer fetched_at: format: date-time type: string performance: format: int64 type: integer seo: format: int64 type: integer title: > PageReviewLighthouse is the page's real Google Lighthouse scores, fetched from type: object types.ArtifactValidationError: properties: column: format: int64 type: integer context: description: | fix the error in place without reading the whole artifact back. title: > Context is a small numbered source snippet around Line so the agent can type: string field: title: | "html" | "js" | "css" | "content" type: string line: format: int64 type: integer message: type: string severity: title: | "error" | "warning" type: string title: > ArtifactValidationError mirrors the internal validator's output for the external API. type: object responses: APIError: content: application/json: schema: externalDocs: url: https://pkg.go.dev/encore.dev/beta/errs#Error properties: code: description: Error code example: not_found externalDocs: url: https://pkg.go.dev/encore.dev/beta/errs#ErrCode type: string details: description: Error details type: object message: description: Error message type: string title: APIError type: object description: Error response securitySchemes: bearerAuth: type: http scheme: bearer description: An Erdo API key (erdo_api_...) or scoped token (erdo_token_...). ````