llms.txt git:20261014.92ad926A

Immutable. This exact content never changes and is served at /api/v1/blob/2da982dda364cbfe.

# Fable-5 for Codex

Fable-5 is an OpenAI Codex plugin for evidence-first audits, deep code review, fact checks, codebase understanding, design options, and repo-wide sweeps.

Latest public tagged release: `v0.4.0-alpha.2`.

```powershell
npx github:rhein1/fable5-codex#v0.4.0-alpha.2
codex plugin add fable5-codex@personal
```

Installed skills:

- `$fable-audit`
- `$fable-deep-review`
- `$fable-fact-check`
- `$fable-understand`
- `$fable-design-options`
- `$fable-sweep`

Fable workflows require source-backed evidence, explicit unknowns, local verification, and a truthful Workflow Trace. Demanding runs keep a GPT-5.6 Sol Ultra coordinator and default bounded subagents to GPT-5.6 Luna medium. They use real subagents only when the runtime exposes them; otherwise they report single-agent multi-lens execution.

Docs:

- README: https://github.com/rhein1/fable5-codex#readme
- Install: https://github.com/rhein1/fable5-codex/blob/main/docs/codex-install.md
- Method: https://github.com/rhein1/fable5-codex/blob/main/docs/method.md
- Development recipes (development branch/main, not the tagged alpha.2 release): https://github.com/rhein1/fable5-codex/blob/main/docs/development-prompts.md
- Optional Laya playbook-selection experiment (source-only, default-off, not the tagged alpha.2 release): https://github.com/rhein1/fable5-codex/blob/main/docs/laya-workflow-selector.md
- Evidence-preserving context packer (source-only; explicit approved inputs, no automatic compaction): https://github.com/rhein1/fable5-codex/blob/main/docs/context-packing.md
- Examples: https://github.com/rhein1/fable5-codex/tree/main/examples/gallery
- Benchmarks and limitations: https://github.com/rhein1/fable5-codex/blob/main/benchmarks/README.md

Boundary: subagents may research and verify, but the main agent owns edits, commits, pushes, reviews, deploys, publication, credentials, and money or wallet actions.