Structured AI development

Jig keeps your AI development work organized, traceable, and verified — specs decompose into scoped tasks, every decision is logged, and verification gates prove the results.

Connect any MCP client
https://mcp.usejig.dev/sse

Add as a remote MCP server in Claude.ai, Claude Code, Cursor, or any MCP-compatible agent.

Or use the CLI
npx jig-dev init

How it works

[1]

Connect

Add the Jig MCP server to Claude.ai, Claude Code, or any MCP-compatible agent. Your agent gets structured project context automatically.

[2]

Spec

Describe what you want built. Jig decomposes it into discrete tasks with acceptance criteria and bounded file scope.

[3]

Verify

Each task must pass verification gates before completion. The agent that writes the code doesn't verify it — Jig does.

What it produces

A bootstrap run generates structured YAML that agents use as ground truth.

.jig/specs/auth.yaml
name: auth
description: JWT authentication with refresh token rotation

architecture:
  pattern: stateless JWT + Redis refresh store
  entry_points:
    - src/auth/middleware.ts
    - src/auth/routes.ts

tasks:
  - id: auth-middleware
    objective: Validate JWT on every protected route
    acceptance_criteria:
      - 401 on missing token
      - 401 on expired token
      - 403 on insufficient scope
    verification:
      - command: bun test tests/auth.test.ts

Why Jig

Organized

Specs decompose into tasks with dependencies. Jig surfaces what's ready and keeps parallel work conflict-free.

Traceable

Every run, every decision, every file touched — logged and reviewable. Requests link work to the reasons it exists.

Scoped

Each task declares exactly which files it may touch. Agents don't wander.

Verified

Verification gates must pass before a task is complete. The agent writing code can't mark its own homework.