Pinned Loading

1. Russian-Judge Russian-Judge Public

   Adversarial AI review that returns a structured verdict: a score, findings classified by severity, and a clear pass/fail floor.

   Python 1

2. CSAE CSAE Public

   Give AI-generated commits a durable audit trail, so you can still prove who approved what, against what scope, months later.

   Shell

3. Docket Docket Public

   A numbered, in-repo work ledger for concurrent AI coding sessions - file it, claim it, close it on the record. Built for Claude Code / multi-agent workflows.

   Python

4. moranbickel moranbickel Public

5. Peer-Worker-Convergence Peer-Worker-Convergence Public

   Run several AI coding sessions on one repo without their branches drifting apart.

   Shell

6. Pre-IMPL-Forensic-Discipline Pre-IMPL-Forensic-Discipline Public

   Check a task's premise against the live codebase before you build, so a stale note doesn't turn into a wrong commit. (v0.1 draft.)