> ## Documentation Index > Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt > Use this file to discover all available pages before exploring further. # Get a task (with steps) ## OpenAPI ````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2026-07/nexus_2026-07.oas.yaml GET /tasks/{id} openapi: 3.0.3 info: title: Nexus API contact: name: Pinecone Support url: https://support.pinecone.io email: support@pinecone.io license: name: Apache 2.0 url: https://www.apache.org/licenses/LICENSE-2.0 version: 2026-07 description: > Nexus turns a set of sources into a queryable, self-improving knowledge base. A context is **sources + a manifest**: upload sources, **curate** them into a searchable index (curate is explicit — there is no auto-curate), and then query the context. The manifest (a single validated JSON document) defines how the context indexes, retrieves, and answers. Conventions: * The `{slug}` path parameter accepts the URL-safe context slug **or** its UUID. * Contexts return a `ContextResponse` (derived flags), not the raw persisted model. Tasks return a `TaskResponse`. * Querying is the unified KnowQL Query API: `POST /api/query` runs one turn (`ask` in, answer out); multi-turn conversations are **sessions** (`GET/DELETE /api/sessions`, `GET /api/queries/{id}`). Read the answer from `output[].content[].text`. * Tasks are a top-level, project-owned resource at `/api/tasks` — they are no longer nested under contexts. Active workflow types are `optimize`, `curate`, `search`, `search_cc`, `search_rag`, `explore`, `profile`, `import`, `pack`, `restore`. servers: - url: https://{host}/api description: Your Nexus deployment host variables: host: default: api.example.com security: - bearerAuth: [] - apiKey: [] tags: - name: Auth description: Login and identity. - name: Project description: The active Pinecone project and its disclosure state. - name: Contexts description: A context is sources + a manifest. Curate it explicitly before querying. - name: Manifest description: Manifest templates and the per-context manifest. - name: Source Files description: >- Per-context source imports (upload, connector, public repo) and the source file tree. - name: Knowledge description: Read-only browse of curated knowledge (chunks + artifacts). - name: Curation description: Curate workflow trigger, curation ledger, and version-pin primitives. - name: Connectors description: >- Project-level links to external source providers (Box, ...). Linked via OAuth or an API key, then used to import source documents into a context. - name: Query description: >- The unified KnowQL Query API: run a query turn (`/query`), fetch a turn (`/queries/{id}`), inspect its trace, manage feedback, comparison flags, and multi-turn sessions (`/sessions`). - name: Tasks description: >- Top-level, project-owned task records for every workflow (optimize, curate, search, work, explore, profile, import, pack, restore, groom). - name: Task Files description: Read, list, and delete files from live or archived task containers. paths: /tasks/{id}: parameters: - $ref: '#/components/parameters/TaskId' - $ref: '#/components/parameters/ApiVersionHeader' get: tags: - Tasks summary: Get a task (with steps) operationId: get_task responses: '200': description: The task content: application/json: schema: $ref: '#/components/schemas/TaskResponse' '404': $ref: '#/components/responses/NotFound' components: parameters: TaskId: name: id in: path required: true schema: type: string description: Task id. ApiVersionHeader: in: header name: X-Pinecone-Api-Version description: >- Date-based contract version. Omit to resolve to the Nexus default version (`2026-07`, the oldest served version); send `unstable` for the in-development surface. The resolved version is echoed back on the same header. A present-but-unrecognized value is rejected with `400 unsupported_api_version`. required: false style: simple schema: type: string default: 2026-07 x-enum: - 2026-07 - unstable schemas: TaskResponse: type: object required: - id - project_id - created_by - workflow - state - input - steps - tokens_prompt - tokens_completion - runtime_seconds - created_at properties: id: type: string project_id: type: string context_id: type: string nullable: true agent_id: type: string nullable: true session_id: type: string nullable: true description: Set for query turns created_by: type: string workflow: $ref: '#/components/schemas/Workflow' state: $ref: '#/components/schemas/TaskState' error: type: string nullable: true input: $ref: '#/components/schemas/TaskInput' output: type: object nullable: true allOf: - $ref: '#/components/schemas/TaskOutput' steps: type: array items: $ref: '#/components/schemas/TaskStep' description: Populated only on GET /tasks/{id} tokens_prompt: type: integer format: int64 tokens_completion: type: integer format: int64 runtime_seconds: type: integer format: int64 running_from: type: string format: date-time nullable: true timeout_seconds: type: integer nullable: true timeout_at: type: string format: date-time nullable: true archived_at: type: string format: date-time nullable: true created_at: type: string format: date-time last_activity_at: type: string format: date-time nullable: true schedule: type: string nullable: true scheduled_at: type: string format: date-time nullable: true parent_task_id: type: string nullable: true progress: $ref: '#/components/schemas/TaskProgress' Workflow: type: string x-enum: - optimize - curate - search - search_cc - search_rag - work - explore - profile - import - pack - restore - groom description: >- The canonical workflow name. Legacy aliases (`build`, `query`, `query_search`, `query_sac`, `query_cc`, `query_rag`, `response`, `agent`, `session`) are still accepted on input and normalized to the canonical value. TaskState: type: string x-enum: - scheduled - starting - provisioning - running - stopping - completed - cancelled - failed TaskInput: description: >- A task's per-workflow `input`, written once at creation and echoed back verbatim. It carries no discriminant of its own — the sibling `workflow` field selects the shape. anyOf: - $ref: '#/components/schemas/SearchInput' - $ref: '#/components/schemas/CurateInput' - $ref: '#/components/schemas/ImportInput' - $ref: '#/components/schemas/PackInput' - $ref: '#/components/schemas/RestoreInput' - $ref: '#/components/schemas/ExploreInput' - $ref: '#/components/schemas/ProfileInput' - $ref: '#/components/schemas/OptimizeInput' - $ref: '#/components/schemas/GroomInput' TaskOutput: description: >- A task's reported `output` — heterogeneous: every workflow reports a different per-state shape, stored and echoed verbatim. An unrecognized or newer shape is preserved as-is rather than rejected, so treat this union as open. anyOf: - $ref: '#/components/schemas/SearchTurnCompletedOutput' - $ref: '#/components/schemas/SearchCcTurnCompletedOutput' - $ref: '#/components/schemas/SearchTurnFailedOutput' - $ref: '#/components/schemas/PackTaskOutput' - $ref: '#/components/schemas/RestoreTaskOutput' - $ref: '#/components/schemas/OptimizeResultOutput' - $ref: '#/components/schemas/OptimizeNoopOutput' - $ref: '#/components/schemas/GroomMetricsOutput' - $ref: '#/components/schemas/CurateRunOutput' - $ref: '#/components/schemas/CurateManifestTuneOutput' - $ref: '#/components/schemas/CurateNoopOutput' - $ref: '#/components/schemas/ExploreOutput' - $ref: '#/components/schemas/ProfileEstimateOutput' - $ref: '#/components/schemas/ProfileEmptyOutput' - $ref: '#/components/schemas/ImportOutput' - $ref: '#/components/schemas/StatusErrorOutput' - $ref: '#/components/schemas/StateFailedOutput' - type: object title: Other additionalProperties: true description: Any shape without a named variant TaskStep: type: object required: - step_id - status description: >- One reported step row. `step_id` and `status` are always set; the rest are per-call optionals. properties: step_id: type: string status: type: string type: type: string commentary: type: string code: type: string description: Redacted tool input (search) result: type: string description: Redacted tool output (search) cum_input_tokens: type: integer format: int64 cum_output_tokens: type: integer format: int64 job_id: type: string description: LlamaParse job id path: type: string description: Source path being parsed TaskProgress: type: object required: - phase - phases - label - pct description: Coarse task progress surfaced to the console. properties: phase: type: integer format: int64 phases: type: integer format: int64 label: type: string pct: type: number format: double eta_seconds: type: integer format: int64 nullable: true Error: type: object required: - message description: The error envelope every non-2xx response uses. properties: message: type: string code: type: string description: >- Stable machine-readable code, present only on errors a client must branch on. x-enum: - unsupported_api_version - preview_not_enabled - preview_file_limit - preview_byte_limit - workspace_host_required - workspace_not_found SearchInput: type: object title: search / search_cc / search_rag / work required: - session_id - query_id - ask - messages - models - tools - shape - instructions - scope - turn_timeout_seconds - retrieval - comparison_group properties: session_id: type: string query_id: type: string ask: type: string description: >- Title fallback for the tasks UI; empty when the turn had no user message messages: type: array items: type: object required: - role - content properties: role: type: string content: type: string models: type: array items: type: string tools: type: array items: type: string shape: type: object nullable: true additionalProperties: true instructions: type: string nullable: true scope: type: array items: type: string turn_timeout_seconds: type: integer format: int64 retrieval: type: object description: >- The turn's retrieval controls; every key is written, `null` where the caller omitted it. required: - retrieval_only - pointers_only - chunks_only - artifacts_only - max_retrieved - max_retrieved_chars - compose - max_steps - thinking_level properties: retrieval_only: type: boolean nullable: true pointers_only: type: boolean nullable: true chunks_only: type: boolean nullable: true artifacts_only: type: boolean nullable: true max_retrieved: type: integer format: int64 nullable: true max_retrieved_chars: type: integer format: int64 nullable: true compose: type: boolean nullable: true max_steps: type: integer format: int64 nullable: true thinking_level: type: string nullable: true x-enum: - minimal - low - medium - high comparison_group: type: string nullable: true CurateInput: type: object title: curate required: - trigger - force - scope properties: trigger: type: string description: e.g. manual / reconcile / restore / resume force: type: boolean scope: type: string description: e.g. full / reconcile / restore batch_size: type: integer models: type: array items: type: string legs: type: array items: type: string description: >- `["chunks"]` gates the run to the chunk + keyword indices, skipping artifact extraction artifacts_reset: type: boolean skip_corpus_aggregation: type: boolean claimed_version: type: integer format: int64 description: The version a drained predecessor had claimed ImportInput: title: import description: >- One of three disjoint shapes, discriminated by which of `connector_id` / `url` / `raw_archive_key` is present. anyOf: - type: object title: Connector import required: - provider - connector_id - item_type - files - path properties: provider: type: string connector_id: type: string folder: type: string nullable: true item_type: type: string x-enum: - folder - file files: type: array items: type: string path: type: string preview: $ref: '#/components/schemas/PreviewLimits' - type: object title: Public-repo import required: - provider - url - path properties: provider: type: string url: type: string path: type: string preview: $ref: '#/components/schemas/PreviewLimits' - type: object title: Archive upload required: - raw_archive_key - path properties: raw_archive_key: type: string description: Staging key under the source bucket's `_inbox/` prefix path: type: string preview: $ref: '#/components/schemas/PreviewLimits' PackInput: type: object title: pack required: - include_traces - slug - context_id properties: include_traces: type: boolean slug: type: string context_id: type: string RestoreInput: type: object title: restore required: - pack_key - pack_filename - target_context_id - target_slug - target_name properties: pack_key: type: string pack_filename: type: string target_context_id: type: string target_slug: type: string target_name: type: string ExploreInput: type: object title: explore required: - guide - trigger properties: guide: type: string trigger: type: string ProfileInput: type: object title: profile required: - manifest - trigger properties: manifest: allOf: - $ref: '#/components/schemas/Manifest' description: The candidate manifest the run profiles, not a stored one. trigger: type: string OptimizeInput: type: object title: optimize description: Every field optional — an input with no keys is a valid no-op optimize. properties: candidate_queries: type: array items: $ref: '#/components/schemas/CandidateQuery' eval_pass_rate_threshold: type: number format: double retrieval_p90_latency_ms: type: integer format: int64 optimize_timeout_seconds: type: integer format: int64 max_tool_turns: type: integer max_iterations: type: integer models: type: array items: type: string GroomInput: type: object title: groom properties: budget: type: integer groom_timeout_seconds: type: integer format: int64 models: type: array items: type: string SearchTurnCompletedOutput: type: object title: search / search_rag completed required: - status - query_id - answer - output_json - citations - latency_ms properties: status: type: string query_id: type: string answer: type: string output_json: type: object nullable: true additionalProperties: true citations: type: array items: type: string latency_ms: type: integer format: int64 SearchCcTurnCompletedOutput: type: object title: search_cc completed required: - status - query_id - answer - citations - latency_ms - steps properties: status: type: string query_id: type: string answer: type: string citations: type: array items: type: string latency_ms: type: integer format: int64 steps: type: integer format: int64 SearchTurnFailedOutput: type: object title: search failed required: - status - query_id - error properties: status: type: string query_id: type: string error: type: string PackTaskOutput: type: object required: - state description: >- A pack task's output, discriminated by `state`. Fields are per-state; absent fields are omitted rather than nulled. properties: state: type: string x-enum: - running - completed - failed - unknown slug: type: string include_traces: type: boolean started_at: type: string completed_at: type: string enriching: type: boolean sources_total: type: integer format: int64 knowledge_total: type: integer format: int64 traces_total: type: integer format: int64 files_written: type: integer format: int64 bytes_collected: type: integer format: int64 pack_filename: type: string pack_size: type: integer format: int64 sources_files: type: integer format: int64 knowledge_files: type: integer format: int64 traces_included: type: boolean traces_files: type: integer format: int64 corpus_groups_included: type: integer format: int64 details: $ref: '#/components/schemas/PackDetails' error: type: string RestoreTaskOutput: type: object title: restore required: - state description: Progressive restore reports, discriminated by which fields are present. properties: state: type: string pack_filename: type: string pack_size: type: integer format: int64 target_context_id: type: string target_slug: type: string stage_progress: type: object required: - phase - done - total properties: phase: type: string done: type: integer format: int64 total: type: integer format: int64 sources_count: type: integer format: int64 knowledge_count: type: integer format: int64 sources_staged: type: integer format: int64 knowledge_staged: type: integer format: int64 manifest_set: type: boolean corpus_groups_set: type: integer format: int64 started_at: type: string completed_at: type: string OptimizeResultOutput: type: object title: optimize result required: - status - best_score - best_iteration - best_latency_ok - total_iterations - pass_rate_threshold - p90_latency_limit_ms - elapsed_seconds - usage - models - manifest - changed_legs properties: status: type: string description: >- `ready` only when the best iteration cleared both the pass-rate and latency bars best_score: type: number format: double best_iteration: type: integer format: int64 best_latency_ok: type: boolean total_iterations: type: integer format: int64 pass_rate_threshold: type: number format: double p90_latency_limit_ms: type: number format: double nullable: true elapsed_seconds: type: number format: double usage: $ref: '#/components/schemas/ChatUsage' models: type: object additionalProperties: type: string description: tier → resolved model id manifest: allOf: - $ref: '#/components/schemas/Manifest' description: >- The tuned manifest the orchestrator persists onto the context when the task completes. changed_legs: type: array items: type: string OptimizeNoopOutput: type: object title: optimize no-op required: - status - reason - candidate_queries - problems description: Too few candidate queries or problems to tune from. properties: status: type: string reason: type: string candidate_queries: type: integer format: int64 problems: type: integer format: int64 GroomMetricsOutput: type: object title: groom required: - status - budget - batches_staged - batches_consumed - facts_distilled - facts_added - facts_merged - facts_dropped_untyped - edges_added - artifacts_pruned - artifact_files_written - artifact_files_deleted - artifact_count - artifact_bytes - duration_ms properties: status: type: string budget: type: integer format: int64 batches_staged: type: integer format: int64 batches_consumed: type: integer format: int64 facts_distilled: type: integer format: int64 facts_added: type: integer format: int64 facts_merged: type: integer format: int64 facts_dropped_untyped: type: integer format: int64 edges_added: type: integer format: int64 artifacts_pruned: type: integer format: int64 artifact_files_written: type: integer format: int64 artifact_files_deleted: type: integer format: int64 artifact_count: type: integer format: int64 artifact_bytes: type: integer format: int64 duration_ms: type: integer format: int64 CurateRunOutput: type: object title: curate run required: - status - flipped - versioned - mode - scope - images_skipped_no_vision - usage - sample_queries - delta properties: status: type: string flipped: type: boolean versioned: type: boolean mode: type: string description: '`incremental` when the context has a manifest, `full` otherwise' scope: type: string images_skipped_no_vision: type: integer format: int64 usage: $ref: '#/components/schemas/CurateUsage' sample_queries: type: array items: type: string description: Example questions the corpus can answer delta: type: object required: - new - changed - deleted - processed - failed - failed_sources - failed_sources_truncated - failed_sources_file - corpus_artifacts - corpus_edges - corpus_artifacts_screened - screened - chunks_created - chunks_updated - artifacts_created - artifacts_updated - chunks_screened - artifacts_screened - images_skipped_no_vision - artifacts_errored - sources_content_truncated properties: new: type: integer format: int64 changed: type: integer format: int64 deleted: type: integer format: int64 processed: type: integer format: int64 failed: type: integer format: int64 failed_sources: type: array items: type: string failed_sources_truncated: type: boolean description: >- Whether failed_sources was capped and the full list spilled to a file failed_sources_file: type: string description: Where the full failed-sources list was written when truncated corpus_artifacts: type: integer format: int64 corpus_edges: type: integer format: int64 corpus_artifacts_screened: type: integer format: int64 screened: type: integer format: int64 description: Sum of the per-leg and corpus screen counts chunks_created: type: integer format: int64 chunks_updated: type: integer format: int64 artifacts_created: type: integer format: int64 artifacts_updated: type: integer format: int64 chunks_screened: type: integer format: int64 artifacts_screened: type: integer format: int64 images_skipped_no_vision: type: integer format: int64 artifacts_errored: type: integer format: int64 sources_content_truncated: type: integer format: int64 CurateManifestTuneOutput: type: object title: curate manifest_tune required: - status - flipped - versioned - mode - usage - delta description: >- The post-optimize artifacts re-eval. Targets optimize trial namespaces, so `flipped` is always false and `mode` is always `manifest_tune`. properties: status: type: string flipped: type: boolean x-enum: - false versioned: type: boolean mode: type: string x-enum: - manifest_tune usage: $ref: '#/components/schemas/CurateUsage' delta: type: object required: - artifacts_reevaluated - failed - failed_sources - failed_sources_truncated - failed_sources_file - corpus_artifacts - corpus_edges - chunks_created - chunks_updated - artifacts_created - artifacts_updated - chunks_screened - artifacts_screened - images_skipped_no_vision - artifacts_errored - sources_content_truncated properties: artifacts_reevaluated: type: integer format: int64 failed: type: integer format: int64 failed_sources: type: array items: type: string failed_sources_truncated: type: boolean failed_sources_file: type: string corpus_artifacts: type: integer format: int64 corpus_edges: type: integer format: int64 chunks_created: type: integer format: int64 chunks_updated: type: integer format: int64 artifacts_created: type: integer format: int64 artifacts_updated: type: integer format: int64 chunks_screened: type: integer format: int64 artifacts_screened: type: integer format: int64 images_skipped_no_vision: type: integer format: int64 artifacts_errored: type: integer format: int64 sources_content_truncated: type: integer format: int64 CurateNoopOutput: type: object title: curate no-op required: - status - flipped - versioned - mode - scope - delta description: >- Nothing changed since the last run; `delta` is the literal string `empty`. properties: status: type: string flipped: type: boolean versioned: type: boolean mode: type: string scope: type: string delta: type: string x-enum: - empty ExploreOutput: type: object title: explore required: - status - matches - none - guide - models - usage properties: status: type: string matches: type: array description: >- Validated template recommendations; unknown or low-confidence proposals are dropped upstream. items: $ref: '#/components/schemas/TemplateMatch' none: type: boolean description: True when no template fit guide: type: string models: type: object additionalProperties: type: string description: tier → resolved model id usage: $ref: '#/components/schemas/ChatUsage' ProfileEstimateOutput: type: object title: profile estimate required: - state - g - density - cost - sampling properties: state: type: string g: type: object required: - value - lo - hi - method - recurrence - confidence - note description: Estimated distinct-artifact count with its uncertainty band. properties: value: type: integer format: int64 lo: type: integer format: int64 hi: type: integer format: int64 method: type: string description: >- The estimator the recurrence blend selected; `empty` on the no-sample path recurrence: type: number format: double confidence: type: string note: type: string density: type: object required: - entities_per_doc - avg_doc_chars description: Measured on the sampling window, not the full curate read-depth. properties: entities_per_doc: type: number format: double avg_doc_chars: type: integer format: int64 cost: type: object required: - by_tier - default_tier properties: by_tier: type: object required: - lite - standard - pro properties: lite: $ref: '#/components/schemas/ProfileTierCost' standard: $ref: '#/components/schemas/ProfileTierCost' pro: $ref: '#/components/schemas/ProfileTierCost' default_tier: type: string description: The tier the context's current `artifact_model` resolves to sampling: type: object required: - n_total - n_sampled - fraction - wall_s description: How much of the corpus the profile actually read. properties: n_total: type: integer format: int64 n_sampled: type: integer format: int64 fraction: type: number format: double wall_s: type: number format: double ProfileEmptyOutput: type: object title: profile empty required: - state - reason properties: state: type: string reason: type: string ImportOutput: type: object title: import required: - status - imported_items - output_files - failed - skipped - skipped_items - manifest_sample - manifest_truncated - manifest_file properties: status: type: string imported_items: type: integer format: int64 output_files: type: integer format: int64 failed: type: integer format: int64 skipped: type: integer format: int64 skipped_items: type: array description: Bounded sample of the skipped records items: $ref: '#/components/schemas/ImportManifestEntry' manifest_sample: type: array description: Bounded sample of the records that produced output items: $ref: '#/components/schemas/ImportManifestEntry' manifest_truncated: type: boolean description: Whether either sample was cut manifest_file: type: string description: Where the full manifest was written manifest_object: type: string description: Object-store key of the full manifest StatusErrorOutput: type: object title: status-keyed failure required: - status - error description: >- The groom / curate terminal-failure convention — best-effort forensics on a task deliberately left Running so the orchestrator can resume it. properties: status: type: string error: type: string StateFailedOutput: type: object title: state-keyed failure required: - state - error description: The pack / restore terminal-failure convention. properties: state: type: string error: type: string PreviewLimits: type: object required: - allow_archives - allowed_extensions - remaining_files - remaining_bytes description: >- Preview-mode import caps grafted onto the input; absent outside preview mode. properties: allow_archives: type: boolean allowed_extensions: type: array items: type: string remaining_files: type: integer format: int64 remaining_bytes: type: integer format: int64 Manifest: type: object description: >- The context manifest that drives curate and search. A sparse document — every field is optional and an omitted key falls back to the validator default, so absent fields are omitted rather than nulled. Value constraints are enforced by the write-side validator. properties: curate: $ref: '#/components/schemas/CurateManifest' optimize: $ref: '#/components/schemas/OptimizeManifest' CandidateQuery: type: object required: - ask description: A query the optimize run tunes toward. properties: ask: type: string answer: type: string description: Used as eval ground truth fallback_to_chunks: type: boolean latency_ms: type: number format: double PackDetails: type: object description: Publisher-facing pack enrichment. properties: description: type: string sample_queries: type: array items: type: string ChatUsage: type: object required: - prompt_tokens - completion_tokens - total_tokens - by_model - cache_read_tokens - cache_write_tokens description: >- The chat-seam token rollup reported by the inference client, which carries the cache-token fields. Used by optimize and explore. properties: prompt_tokens: type: integer format: int64 completion_tokens: type: integer format: int64 total_tokens: type: integer format: int64 by_model: type: object description: >- Keyed by the model the proxy actually dispatched to, so tier fallbacks are visible. additionalProperties: $ref: '#/components/schemas/ModelUsage' cache_read_tokens: type: integer format: int64 cache_write_tokens: type: integer format: int64 CurateUsage: type: object required: - prompt_tokens - completion_tokens - total_tokens - by_model - embed_by_model description: >- Curate's token rollup: the chat-seam totals plus the embed/rerank seam (`embed_by_model`). It does not carry the cache-token fields. properties: prompt_tokens: type: integer format: int64 completion_tokens: type: integer format: int64 total_tokens: type: integer format: int64 by_model: type: object additionalProperties: $ref: '#/components/schemas/ModelUsage' embed_by_model: type: object additionalProperties: $ref: '#/components/schemas/SeamUsage' TemplateMatch: type: object required: - template_id - rationale - confidence properties: template_id: type: string rationale: type: string confidence: type: number format: double ProfileTierCost: type: object required: - tokens - time_s properties: tokens: type: object required: - point - lo - hi - input - output description: >- `point` is the estimate at G's midpoint and splits into input/output; `lo`/`hi` are band totals. properties: point: type: integer format: int64 lo: type: integer format: int64 hi: type: integer format: int64 input: type: integer format: int64 output: type: integer format: int64 time_s: type: object required: - lo - hi properties: lo: type: integer format: int64 hi: type: integer format: int64 ImportManifestEntry: type: object required: - src - action - out - fmt - reason - bytes description: One ingest record. properties: src: type: string action: type: string x-enum: - parsed - copied - extracted - skipped - failed out: type: array items: type: string description: Empty exactly when nothing was written fmt: type: string reason: type: string bytes: type: integer format: int64 CurateManifest: type: object properties: chunks: $ref: '#/components/schemas/ChunksConfig' artifacts: $ref: '#/components/schemas/ArtifactsConfig' OptimizeManifest: type: object description: Scheduled self-tuning config. properties: schedule: type: string maxLength: 128 default: 0 * * * * description: cron latency_threshold_ms: type: integer minimum: 0 maximum: 600000 default: 60000 min_group_size: type: integer minimum: 1 maximum: 100 default: 2 eval_pass_rate_threshold: type: number format: double minimum: 0 maximum: 1 default: 1 max_iterations: type: integer minimum: 1 maximum: 100 default: 20 ModelUsage: type: object required: - tier - prompt_tokens - completion_tokens - cache_read_tokens - cache_write_tokens - total_tokens description: Per-model chat-seam counts; both usage roll-ups write every key. properties: tier: type: string prompt_tokens: type: integer format: int64 completion_tokens: type: integer format: int64 cache_read_tokens: type: integer format: int64 cache_write_tokens: type: integer format: int64 total_tokens: type: integer format: int64 SeamUsage: type: object required: - kind - tier - embed_tokens - rerank_units description: >- Per-model embed/rerank-seam counts: billed per token (embed) or per request (rerank). properties: kind: type: string tier: type: string embed_tokens: type: integer format: int64 rerank_units: type: integer format: int64 ChunksConfig: type: object description: Source chunking + embedding config. properties: enabled: type: boolean default: true embedding_model: type: string maxLength: 128 default: multilingual-e5-large chunking: $ref: '#/components/schemas/ChunkingConfig' keyword: type: object properties: enabled: type: boolean default: true ArtifactsConfig: type: object description: Knowledge-artifact extraction config. properties: enabled: type: boolean default: false artifact_model: type: string x-enum: - standard - lite default: lite artifact_types: type: array maxItems: 32 items: $ref: '#/components/schemas/ArtifactType' edge_types: type: array maxItems: 32 items: $ref: '#/components/schemas/EdgeType' min_doc_count: type: integer minimum: 1 maximum: 64 default: 1 max_tokens: type: integer minimum: 128 maximum: 8192 default: 1500 max_doc_chars: type: integer minimum: 1000 maximum: 2000000 default: 60000 extraction_window_chars: type: integer minimum: 0 maximum: 1000000 default: 0 mention_max_chars: type: integer minimum: 100 maximum: 4000 default: 400 max_mentions_per_artifact: type: integer minimum: 1 maximum: 500 default: 40 mention_context_chars: type: integer minimum: 1000 maximum: 100000 default: 8000 max_artifacts_per_type: type: integer minimum: 1 maximum: 10000 default: 10000 ChunkingConfig: type: object properties: strategy: type: string x-enum: - markdown_heading - sliding_window - html - csv - code default: markdown_heading target_size: type: integer minimum: 64 maximum: 4096 default: 512 overlap: type: integer minimum: 0 maximum: 1024 default: 64 respect_sections: type: boolean default: true ArtifactType: type: object required: - name properties: name: type: string maxLength: 64 kind: type: string x-enum: - summary - topic - entity - event - doc - page - glossary default: topic scope: type: string x-enum: - document - corpus default: corpus icon: type: string description: type: string coverage: type: array maxItems: 32 items: type: string maxLength: 128 sections: type: array maxItems: 64 items: type: string maxLength: 128 min_doc_count: type: integer minimum: 1 maximum: 64 default: 1 format: type: string x-enum: - markdown - sqlite default: markdown columns: type: array maxItems: 64 description: sqlite-format artifacts only items: type: object required: - name properties: name: type: string maxLength: 64 type: type: string x-enum: - TEXT - INTEGER - REAL - NUMERIC default: TEXT description: type: string natural_key: type: array maxItems: 16 items: type: string maxLength: 64 EdgeType: type: object required: - name - from - to properties: name: type: string maxLength: 64 from: type: string maxLength: 64 description: Source artifact-type name to: type: string maxLength: 64 description: Target artifact-type name description: type: string attributes: type: array maxItems: 16 items: type: string maxLength: 64 responses: NotFound: description: Not found content: application/json: schema: $ref: '#/components/schemas/Error' securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: JWT description: >- Session token from POST /auth/login. Pass as `Authorization: Bearer `. The alternative `X-Pinecone-Api-Key` header is also accepted for direct-key auth (used by the Nexus CLI on first contact). apiKey: type: apiKey in: header name: X-Pinecone-Api-Key description: Pinecone API key, accepted as an alternative to the bearer JWT. ````