Use this section for threat models, disclosure policy details, security architecture notes, and hardening guidance.
| Doc | Topic |
|---|---|
| llm-data-flow.md | What data leaves the cluster → OpenAI on each judgment generation (feat_llm_judgments) |
| github-token-handling.md | Per-repo PAT storage / rotation / scopes / leak-prevention (feat_github_pr_worker) |