Player profile
GraphCode
A native macOS workspace that arranges coding-agent sessions into a graph, where every node is a live terminal and edges fire hand-offs, messages, and spawns without a person present.
Profile
Where it fits
GraphCode is a native macOS app that runs each coding-agent session as a node on a canvas. Nodes are real terminals backed by libghostty and a zmx multiplexer, so a session can be opened, watched, and corrected mid-run instead of only inspected after it finishes.
It belongs in Parallel Coding-Agent Runners because coordination is expressed as typed edges between sessions: hand-off, message, and spawn edges fire unattended, loops can be goal-based (run until a shell command exits 0), and a local daemon keeps sessions alive across app restarts. Claude Code is the most complete backend, with Codex and Copilot CLI also supported.
Builder web analytics
Measure projects built with GraphCode
GraphCode can run a graph of coding-agent sessions that ship changes to a product surface unattended. Agent Analytics measures whether the work those edges triggered actually improved the user-facing result.
Instrument the deployed surface that GraphCode loops change. Agent Analytics reads product and web events after the work lands; it does not replace GraphCode's session transcripts, dial log, graph state, or agent output.
First loop to measure
- a GraphCode loop asks Claude Code, Codex, or Copilot CLI to change a website, docs path, onboarding flow, app surface, or experiment
- the changed surface reports visits, sources, CTA clicks, signup, activation, retention, or task-completion events to Agent Analytics
- a hand-off or message edge fires a follow-up loop that fetches Agent Analytics results after the change ships
- the next goal-based loop is scoped from measured user outcomes rather than from a shell predicate alone
Copyable prompt
Use Agent Analytics for this project. If event reporting is missing, add the tracker and report events for this project surface, including GraphCode-managed page, docs path, traffic source, CTA click, signup, activation event, retention signal, or shipped experiment. Verify events are arriving. Then fetch the last 7 days and compare them with the prior 7 days. Tell me which GraphCode-managed page, docs path, traffic source, CTA click, signup, activation event, retention signal, or shipped experiment moved users toward value, where users dropped off, which sources mattered, and what my agent workflow should improve next.