
fable-thinking
Reasoning protocol distilled from Claude Fable 5, packaged as a Claude Code plugin — evidence-grounded, calibrated, hard to fool.
Install with your AI
Paste into Claude Code, Cursor, or any agent — it reads the repo and wires the tool into your project.
Install and set up fable-thinking (claude-plugin project) into my current project. Found on https://claudeers.com/fable-thinking Repo: https://github.com/mrgoonie/fable-thinking Homepage/docs: — Detected install method: claude-plugin → /plugin install fable-thinking@mrgoonie/fable-thinking Category: plugins. Platforms: api, web. Read the repo's README for exact setup and env vars, then install it and wire it into my project. Claudeers Health Verdict: active; community-verified: false. Confirm the source before running anything.
/plugin marketplace add mrgoonie/fable-thinking /plugin install fable-thinking@mrgoonie/fable-thinking
git clone https://github.com/mrgoonie/fable-thinking
// compatibility
| Platforms | api, web |
|---|---|
| Operating systems | — |
| AI compatibility | claude |
| License | MIT |
| Pricing | open-source |
| Language | HTML |
Fable Thinking
The reasoning discipline of Claude Fable 5, distilled into an executable protocol — packaged as a Claude Code plugin.
→ View the live landing page · dark/light theme · EN / VI
Fable Thinking is not a persona to imitate. It is a set of mechanical procedures that make any model's reasoning more grounded, better calibrated, and harder to fool — including by its own fluent output. It cannot add capability; it removes the predictable failure modes that waste whatever capability the executing model has.
It works across models and runtimes (Claude, Codex/GPT, Gemini, local models).
What it does
- The Floor — three fast checks run before every answer (even simple-looking ones), catching confident template answers to trick questions.
- Multi-hypothesis diagnosis — forces discriminating tests instead of pattern-match satisfaction.
- Concrete simulation — runs the movie of what actually happens before committing to an answer.
- Adversarial self-review — treats your own fluent prose as a suspect, not proof.
- Constraint Loop — mechanically verifies hard output constraints (banned letters, exact counts, strict formats) unit by unit, instead of rubber-stamping them.
- Calibrated, outcome-first delivery — states claims with the confidence the evidence supports.
When to use it
Invoke when a task needs careful reasoning rather than a routine answer:
- Debugging and root-cause analysis
- Code review and correctness checks
- Architecture and strategy decisions
- Contested or high-stakes claims
- Constrained writing (letter bans, word counts, acrostics, strict formats)
- Any task where being right matters more than being fast
Installation (Claude Code)
This repository is a Claude Code plugin marketplace. Add it, then install the plugin:
/plugin marketplace add mrgoonie/fable-thinking
/plugin install fable-thinking@fable-thinking
Or add the marketplace from its Git URL:
/plugin marketplace add https://github.com/mrgoonie/fable-thinking
After installing, the fable-thinking skill becomes available to Claude Code and is invoked automatically when a task calls for careful reasoning, or explicitly by name.
Repository layout
fable-thinking/
├── .claude-plugin/
│ ├── marketplace.json # Marketplace definition
│ └── plugin.json # Plugin manifest
├── skills/
│ └── fable-thinking/
│ ├── SKILL.md # The reasoning protocol
│ └── references/ # Content, design, and worked examples
├── web/ # Landing page (published to AgentWiki)
│ ├── index.html
│ └── assets/hero-dark.png
└── README.md
Live page
The landing page is published at api.agentwiki.cc/s/4pscSZumUfvoLT6jUUxz7 — dark/light theme, English and Vietnamese.
Credits
Distilled from the reasoning discipline of Claude Fable 5. Originally authored as part of AgentKit.
License
// faq
What is fable-thinking?
Reasoning protocol distilled from Claude Fable 5, packaged as a Claude Code plugin — evidence-grounded, calibrated, hard to fool.. It is open-source on GitHub.
Is fable-thinking free to use?
fable-thinking is open-source under the MIT license, so it is free to use.
What category does fable-thinking belong to?
fable-thinking is listed under plugins in the Claudeers registry of Claude-compatible tools.
// embed badge
[](https://claudeers.com/fable-thinking)
// retro hit counter
[](https://claudeers.com/fable-thinking)
// reviews
// guestbook
// related in Claude Plugins
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explainin…
"CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/
financial-services — a Claude ecosystem project on GitHub.
