Now in Beta — Limited Access

The missing context layer
for AI coding agents.

Carrick maps your APIs, types, and dependencies across every repo, so your agent knows what's on the other side before it writes a single line.

Runs in CI. Queryable via MCP.

feature/update-orders → main

↓

GitHub Actions · carrick / analyze

See it work

3 repos. 90 seconds. Zero guesswork.

▶

Demo video coming soon

Function Exploration

Every function. Every service. Fully resolved.

Your agent discovers what exists across services, gets resolved types and call graphs, and checks compatibility — all from structured data. It can write a correctly-typed API call without reading a single line of source.

⟠carrick list_function_intents

◆user-service

Discover

Map your entire service topology

"What services exist?"

Service catalog with endpoint counts and type availability

"What functions does this service have?"

Exported functions with intent descriptions and call graphs

Understand

Write a typed API call from a single query

"What does this function do, and what types does it use?"

Route, resolved types, external calls, intent — enough to write a typed API call

"Show me the implementation"

Commit-pinned line coordinates to fetch just the function body

Validate

Catch breaking changes before they ship

"Will this change break anything?"

Missing endpoints, type mismatches between consumer and producer

"What dependencies does this service use?"

Package versions and org-wide version conflicts

Live System Map

See how your services connect.

See all your services across your organisation. A service map is generated in every PR.

CI Integration

Type errors caught at the boundary.

Carrick runs in your GitHub Actions pipeline. When a producer and consumer disagree on types, you know before it ships.

github-actionsbot
🪢 CARRICK: API Analysis Results
Analyzed 21 endpoints and 11 API calls across all repositories.
Found 4 critical mismatches, 22 connectivity issues, 1 dependency conflict.
Type Compatibility Issue: GET /users/:id/comments
Producer: { userId: number; comments: Comment[]; }
Consumer: Comment[]
Error: missing properties from Comment[]: length, pop, push, concat, and 28 more.
Type Compatibility Issue: GET /api/comments (3 occurrences)
Producer: Comment[]
Consumer: { id: string; order_id: string; }[]
Error: Comment[] not assignable to { id: string; order_id: string; }[]

Blog

Notes from the boundary.

Deep Dive

The Multi-Repository TypeScript Problem

Navigating Type Safety Across Service Boundaries

Coming Soon

The Future of CI

Why tests are the wrong abstraction for agents

FAQ

Common questions.

Carrick runs as a GitHub Action in your CI pipeline. Powered by SWC — the compiler behind Vite and Turbopack — it statically analyses your code across repositories, mapping API endpoints, resolving TypeScript types, and building a dependency graph. The results are stored as structured metadata and made available to AI coding agents via MCP (Model Context Protocol).

Carrick stores only API metadata — endpoint signatures, resolved types, and dependency information. It never stores or transmits your source code. Your organisation's data is isolated using unique API keys, and all storage is encrypted at rest.

Currently JavaScript and TypeScript. Once the core tooling is refined, we plan to expand to other major languages.

We'll be rolling out a generous free tier alongside paid plans in line with other CI tooling. Beta users will get early access to pricing details.

Beta members get access to a private Discord server with dedicated channels for support, feature requests, and discussion. Invitations are sent to approved beta users via email.

The missing context layerfor AI coding agents.