Player profile
Agent 007
A self-hosted job board and pixel-art office for running Claude Code, Codex, and other terminal coding agents in parallel, where each queued job gets its own git worktree and branch and ends as a pull request for review.
Profile
Where it fits
Agent 007 is an MIT-licensed Node.js server and browser UI that runs existing coding-agent CLIs, such as Claude Code, Codex, and Gemini CLI, as real PTY sessions. Jobs posted to its board are dispatched to fresh agents, each in its own git worktree branched from the remote base, and an agent finishes a job by handing over the pull request it opened or a written summary. Cards can also run on a cron schedule.
It belongs in Parallel Coding-Agent Runners because several agents work on the same repository at once without sharing a checkout. An optional coordinating agent can plan work, post cards, review results, and answer workers' permission requests; agents can message each other and post jobs over MCP; and a pixel-art office view shows which agents are working and which are waiting on a person.
Builder web analytics
Measure projects built with Agent 007
Agent 007 tracks whether a job produced a reviewed pull request; it does not observe what the merged change did for a product's users. Agent Analytics can measure that for a separately instrumented project.
This is an optional measurement workflow for projects built with Agent 007, not a built-in Agent 007 integration or a claim that its sessions report to Agent Analytics. Configure and verify the project's analytics separately.
First loop to measure
- a job card asks an agent to change a user-facing page or flow
- the agent opens a pull request and a person reviews and merges it
- the deployed project reports page views, signups, and activation events to Agent Analytics
- a follow-up job card compares the observed metrics with the prior period and proposes the next change
Copyable prompt
Use Agent Analytics for this project. If event reporting is missing, add the tracker and report events for this project surface, including page visit, traffic source, signup, activation event, or funnel step. Verify events are arriving. Then fetch the last 7 days and compare them with the prior 7 days. Tell me which page visit, traffic source, signup, activation event, or funnel step moved users toward value, where users dropped off, which sources mattered, and what my agent workflow should improve next.