API reference
DocuClipper REST API for extracting structured data from bank statements, invoices, and tax forms.
Last updated
The DocuClipper API documentation has moved.
→ API docs, getting started, authentication, and endpoint reference.
Quick links
- Getting started, PAT auth, presigned upload, your first extraction.
- Authentication, Personal Access Tokens (
dcp_…) for the Agent API. - For AI agents, End-to-end cURL example, MCP tool dispatch.
- OpenAPI spec, Machine-readable schema (also at
/api-docs/.well-known/openapi.json). - llms.txt, Compact reference for LLM-based clients.
Endpoint overview
Recommended (Agent API, /api/v1/agent/*): PAT auth, direct-to-S3 upload, clean JSON output.
| Method | Endpoint | Purpose |
|---|---|---|
| POST | /agent/documents/upload-url | Get a presigned S3 upload URL |
| POST | /agent/jobs | Create an extraction job |
| GET | /agent/projects | List the projects you can read |
| GET | /agent/projects/:projectId/documents | List the documents in a project |
| GET | /agent/documents/:documentId | Document status and summary |
| GET | /agent/documents/:documentId/transactions | Read extracted transactions (JSON or CSV) |
| GET | /agent/documents/:documentId/data | Structured payload for one document |
| GET | /agent/whoami | Verify your PAT |
| GET/POST | /agent/mcp/tools | List or execute MCP tools |
Extracted data is stored per document, so the /agent/documents/* endpoints are the direct read. The job-keyed equivalents (/agent/jobs/:id, /agent/jobs/:id/transactions, /agent/jobs/:id/data) remain supported for existing integrations, but they can only reach jobs your own token created. Use /agent/projects/:projectId/documents to find documents that arrived through a connected cloud folder, an ingestion email address, or the web UI.
Legacy (/api/v1/protected/*): still supported for existing customers, see the per-document-type guides under /api-docs/.