Referencia de API

Create an analysis

Analyze a contract from text, URL, or PDF (multipart/form-data). Returns 202 immediately with an ana_* handle to poll. Test keys run a deterministic mock engine; live keys run the full LLM pipeline.

{ sourceType, text?, url?, title?, forceReanalysis? } or multipart { file, title? }

{ analysisId: "ana_*", status: "queued", links: { self, result } }

List analyses

List analyses for the authenticated API key, newest first. Results are isolated by organization and mode (test keys never see live analyses).

{ data: AnalysisListItem[], nextCursor? }

Get analysis status and metadata

Returns current status, sourceType, mode, and executionMode. Poll this until status is "completed" or "failed", then fetch the result.

{ analysisId, status, title?, sourceType, mode, executionMode, createdAt }

Get structured analysis result

Returns the full analysis output: summary, risk score, risks, obligations, key dates, and evidence. Only available when status is "completed".

{ analysisId, status, summary, riskScore, risks[], obligations[], keyDates[], evidence[], mock? }

Save an analysis as a contract

Persists a completed analysis (ana_*) as a real contract (ctr_*) with version tracking. Idempotent: re-submitting the same analysisId returns the existing contract with 200.

{ analysisId: "ana_*" }

{ contractId: "ctr_*", versionNumber: 1, analysisId, links: { self } }

List contracts

List saved contracts for the authenticated API key, newest first. Results are isolated by organization and mode.

{ data: ContractListItem[], nextCursor? }

Get a contract

Returns contract details including title, category, risk score, and the current version's analysisId.

{ contractId, title?, status, riskScore?, currentVersion: { versionNumber, analysisId }, createdAt }

Compare two contracts or analyses

Compares two sides synchronously. Each side is either a contract handle (ctr_*, optionally a versionNumber) or an analysis handle (ana_*). Both sides must be completed and share the same mode. Returns cmp_* with a resolved field showing exactly which versions were compared.

{ left: { contractId? | analysisId?, versionNumber? }, right: { contractId? | analysisId?, versionNumber? } }

{ comparisonId: "cmp_*", status, resolved: { left, right }, links: { self } }

Get a comparison

Returns the comparison result including summary, score, recommendation, per-dimension breakdown, and the resolved sides (which analysis versions were actually compared).

{ comparisonId, status, executionMode, summary, score, dimensions[], resolved: { left, right }, createdAt }