First Run

Run a writer and reviewer on a small coding task. Claude writes the function and its test. Node runs the test, then Codex reviews the result. Set up your project as Installation describes.

Run the team

You need the Claude Code and Codex command line tools installed and signed in. A Claude seat writes a function and its test from the brief at briefs/add.md, Node runs the test, and a Codex seat reviews the result. The file and its brief are on A writer and a reviewer. Run it from the directory the work belongs in, or point it at one (where a workflow lives):

Terminal

npx tsx writer-reviewer-pair.ts

Output

{
  "status": "pass",
  "summary": "dag \"writer-reviewer-pair\": all 3 node(s) green",
  "data": {
    "write": {
      "status": "pass",
      "summary": "Created src/add.mjs with a pure add function and test/add.test.mjs with Node.js test coverage"
    },
    "test": {
      "status": "pass",
      "summary": "`node` exited 0"
    },
    "review": {
      "status": "pass",
      "summary": "Review panel: 1/1 reviewer(s) cleared.",
      "data": {
        "findings": [],
        "escalatedFindings": [],
        "errors": [],
        "results": [
          {
            "kind": "verdict",
            "name": "review-1",
            "met": true,
            "reason": "The named add(a, b) function returns a + b without side effects, and the required Node test passes."
          }
        ],
        "passed": 1,
        "required": 1,
        "severityCounts": {}
      }
    }
  }
}

All three stages passed: the writer left src/add.mjs and test/add.test.mjs behind, the test exited 0, and the one reviewer accepted the change. Change the brief, the file list and the test command, and it’s your own work. The file is the short form, workflow(): a list of stages. When a step needs your own code, a branch or a team inside a team, the same team is a dag(), and Workflows shows both forms side by side.

Next steps

Workflows: how the steps, reviews and decisions in the workflow fit together.
Feedback loops: what happens when a review or a test says no.
Examples: complete teams by field, with what each run printed.

Start

Concepts

Patterns

Examples

Evals and Gates

Recording Runs

Driving Runs

Surfaces and Hosts

Writing Workflow Shapes

Memory

Workspace

Compare

FAQ

API Reference

Run the team

Next steps

​Run the team

​Next steps

Run the team

Next steps