Every piece of engineering intelligence, via API
RESTful API with full access to teams, contributors, sprints, code quality, risk scores, reports, and alerts. Build custom integrations, embed insights into internal tools, or automate workflows that Gitrevio doesn't cover out of the box.
Start with one curl command
FastAPI backend, OpenAPI 3.1 auto-generated at /api/v1/docs.
Import the spec into Postman, Insomnia, or generate clients in any language.
SHA-256-hashed API keys with persona-bound scoping. A team lead's key returns their teams; a CTO's key returns the org. Role enforced at the database layer via per-customer Postgres reader role — not application-level filtering.
RFC 7807 problem-details error responses. Cursor-based pagination on every list endpoint. Per-key rate limiting with bucket state visible in response headers.
Mock-first framework. Every endpoint has a mock provider — develop integrations against deterministic fixtures before touching real data.
Endpoint coverage
Seven Layer-1 endpoints ship today. Layer-2 (skill-execution endpoints, write actions, scheduling) is rolled out incrementally — each behind a versioned route with a mock provider.
Coming through Q3–Q4 2026: dedicated routes for contributors, knowledge graph, what-if, release risk, and persona dashboards. Until those ship, the same data is available through /v1/skills/{name}/run — every skill exposes the same canonical data, structured.
Webhooks for real-time events
Don't poll — subscribe. Gitrevio sends webhook events when something changes, so your integrations can react in real time.
Available webhook events: alerts triggered, reports generated, health score changes, sprint status updates, attrition risk changes, and new contributor onboarding milestones.
Every payload is signed with HMAC-SHA256 so you can verify authenticity. Retry logic with exponential backoff handles transient failures on your end.
Use webhooks to power custom Slack bots, update internal dashboards, trigger CI/CD pipeline adjustments, or feed data into your own analytics stack.
API for machines. MCP for AI.
The REST API is the right choice for traditional integrations — scripts, dashboards, internal tools, data pipelines. Structured inputs, structured outputs, predictable behavior.
For AI-native access, pair the API with Gitrevio's MCP server. Your AI agents get the same data through a protocol designed for language models — with tools, prompts, and structured responses they can compose and reason about.
Same data, same RBAC, two access patterns. Use whichever fits your integration.
SDKs and tooling
OpenAPI 3.1 is shipped today. Auto-generate clients in any language with
openapi-generator. Import directly into Postman or
Insomnia. The MCP server is itself a thin client over the same spec.
Official Node.js and Python SDKs are queued for Q1 2027. Until then, the generated clients work — and the MCP server is the simplest path for AI-agent use.