Your OpenAPI spec and your documentation drift apart with every release. DocsCI diffs them on every PR — catching renamed parameters, removed endpoints, and schema mismatches before developers encounter them.
Per-field
Diff granularity
OpenAPI 3.x
Spec format
SARIF
Export format
Every PR
Check frequency
The spec gets updated first. Or the docs get updated first. Either way, drift accumulates. Parameter names change, response fields are added or removed, authentication schemes evolve — and the docs lag behind.
DocsCI imports your OpenAPI spec (from a URL or file path) and compares it against every documented endpoint, parameter, and response schema in your Markdown, MDX, or RST docs. On every pull request, it reports what drifted — specific fields, line numbers, and AI-generated suggestions for how to fix the docs.
For every documented API endpoint, DocsCI checks: does the spec agree? Mismatched paths, methods, params, and schemas are reported as findings with severity.
Runs on every PR that touches docs or the spec file. Catches drift the moment it's introduced — not after it ships to users.
DocsCI validates field names, types, required/optional, enum values, and descriptions — not just endpoint existence. Granular accuracy at the field level.
When a drift finding is reported, Claude generates the corrected docs snippet — ready to apply as a patch diff in your PR.
Export findings as SARIF 2.1.0 for GitHub Code Scanning. Drift findings appear as code annotations in the GitHub UI alongside other CI checks.
Point at your spec with openapi.url in docsci.yml. Supports OpenAPI 3.x from a URL or a file path. Works with Swagger 2.0 via conversion.
Configure the spec URL or path in docsci.yml. DocsCI fetches it on every run.
# docsci.yml openapi: url: https://api.example.com/openapi.json # or path: ./openapi.yaml checks: drift: true
DocsCI parses your documentation, extracts API references, and diffs against the spec. Every divergence is a finding.
Drift findings appear as PR comments with exact line numbers and AI-suggested doc updates. Block the merge until docs match the spec.
Which OpenAPI versions does DocsCI support?
DocsCI supports OpenAPI 3.0.x and 3.1.x natively. Swagger 2.0 specs are automatically converted during import.
How does DocsCI parse documented endpoints from Markdown?
DocsCI uses a combination of fenced code block parsing (for request/response examples) and heuristic Markdown parsing (for parameter tables, path descriptions). It recognizes common documentation patterns from Docusaurus, Mintlify, Nextra, and hand-written Markdown.
Can I configure which drift types are blocking vs advisory?
Yes. In docsci.yml, configure checks.drift_blocking: [renamed_field, removed_endpoint] to make specific drift types block the PR. Others are reported as warnings.
Does DocsCI validate request body schemas?
Yes. DocsCI extracts JSON request body examples from code blocks and validates them against the spec's requestBody schema. Missing required fields and type mismatches are reported.
Join API and platform teams who use DocsCI to ship verified documentation.