Projects / Case study

cstack

Agent skills for building and fixing code. Before anything ships, a Claude and a Codex adversary try to break it, and the loop only ends when a fresh pair finds nothing.

Role
Solo builder
Timeline
November 2026 to now
Built with
Markdown skills, Bash, Python, gh
Code
Public on GitHub, MIT

The problem

Coding agents are quick, but they mark their own homework. An agent that wrote a change is the worst reviewer of it: it shares every assumption it made while writing it.

I was also doing the same steps by hand on every task. Agree what done looks like, build it, ask another model to review it, check the review was right, open a PR, watch CI. I wanted that written down once, so every agent on every machine follows it.

What I built

cstack is a set of skills that Claude Code and Codex load by name. One command installs them on a machine, and one repo serves every machine I use.

  1. GoalThe agent writes the goal, the problem and the smallest change it expects, then waits for my OK. After that the goal is frozen.
  2. Build or fixcstack-build learns the code, builds, and proves it works in the app. cstack-fix reproduces the bug, proves the cause, commits a failing test, then fixes it.
  3. Break loopTwo fresh adversaries, one Claude and one Codex, get the same brief and try to break the change. Every finding needs a repro.
  4. PRfull-stop files one PR with every round's findings, watches CI and review bots, and stops when it's ready to merge. Say "full send" and it merges once the checks pass.
# what I actually type cstack-fix users get logged out after 5 minutes cstack-build add a dark mode toggle to settings, full send full stop

How the break loop works

The agent that wrote the change is the developer. It never judges its own work alone. It reruns every repro the adversaries send. A finding that reproduces gets fixed, one commit each, or goes to an arbiter from the other model family. The developer can't drop a finding or call it out of scope by itself.

Then a new pair of adversaries, who know nothing about earlier rounds, gets the identical brief. The loop ends when a fresh round finds nothing real. There's no round limit.

The steps are the real ones from the skills. Opus is the Claude adversary, Sol the Codex one, and the arbiter always comes from the family that didn't write the code. Timings are illustrative, not from a logged run.

What I borrowed

full-stop, full-send and parts of the break loop are adapted from Theo's tstack. A small file records the exact upstream commit for each one, and cstack upstream shows what changed there since I copied it.

What's next