oh-my-fable plugin
A Claude Code plugin that injects Anthropic guide rules into sessions, with the fable command expanding a vague ask
- 61 ★ on the repository
- Last commit
Extended access
Why this level
- A hook injects rules into every session and shapes agent behavior, and `/fable` runs the request it builds.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/oh-my-fableDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- A hook injects rules into every session and shapes agent behavior, and `/fable` runs the request it builds.
This is third-party code. Review the repository files before installing.
What it does
`oh-my-fable` brings rules from Anthropic's official prompting guides into Claude Code: autonomy, scope limits, delivering what was asked, targeted edits, progress updates and batched tool calls. A hook adds them at the start of every session, and unattended sessions (`claude -p`, the Agent SDK) get separate paragraphs. Subagents receive a short version. The `/fable` command takes a short ask such as "fix the login button" and shows a request with goal, context, scope and done criteria, then runs it. Three more skills: `/fable-status` shows the rules in effect, `/fable-audit` checks `CLAUDE.md` against the guide and `/fable-setup` changes settings. The plugin creates no project files.
Who it is for. Claude Code users who write short requests and want fewer clarifying questions, stray edits and half-finished runs.
Good fit when
- A one-line request makes the agent ask back or edit extra files
- Working rules must apply in every session with no manual `CLAUDE.md` entries
- You want to know what your own rules are missing with `/fable-audit`
- Claude runs unattended through `claude -p` or the Agent SDK
Not a fit when
- Claude Code is older than 2.1.258, which the plugin requires as a minimum
- You have strict rules of your own that conflict with the Anthropic guide
- You need a plugin for Cursor or Codex
Example request
/fable fix this: the login button does nothing after yesterday's auth change.Limitations
It needs Claude Code 2.1.258 or newer, and on Windows also Git for Windows (Git Bash). Rules are injected in English verbatim from the guides and tuned for Opus 5.5, Fable 5.2 and 5.1; other models behave the same, with differences documented. The request card from `/fable` is drawn in the terminal from version 2.1.287 and in the desktop app. CI and cloud sessions need separate setup from the environments document.
How to disable. Run `/fable-setup remove`, then `claude plugin uninstall oh-my-fable@oh-my-fable`. By default the plugin leaves nothing in the project.
FAQ
How do I install it?
Tell Claude Code to install https://github.com/Junhan2/oh-my-fable, or run two commands: `claude plugin marketplace add Junhan2/oh-my-fable` and `claude plugin install oh-my-fable@oh-my-fable`.
What does `/fable` do?
It turns a short ask into a request with goal, context, scope and done criteria and runs it. With "just the prompt" it only shows the request.
Does it change `CLAUDE.md`?
No, with the default delivery `CLAUDE.md` and rules files stay untouched, so nothing remains after uninstalling.
How do I audit my own rules?
With `/fable-audit`, which reads `CLAUDE.md` and rules files, checks them against the guide and changes nothing.
Does it work on other models?
Yes, the README lists Sonnet 5.5, Haiku 5.5, Opus 5 and Sonnet 5 with the same behavior. Per-model differences are collected in the How it works page.
Discussion
No questions or reviews yet.
Related
PluginsA Claude Code plugin with four working-method skills: classify the task, act in steps, verify the result and generate domain adapters
A self-improvement plugin for Claude Code, Codex and OpenCode: turns corrections into rules the agent follows in later sessions
An execution-discipline skill for Claude: staged planning, delegation to agents, failable checks and a double-check before delivery
claude-mods adds 112 skills, 3 agents and 3 commands to Claude Code: stacks, agent operation, defense, session continuity.
A Claude Code plugin that makes the model finish the job: result verification, evidence and systematic investigation enforced as procedure
Multi-agent orchestration for Claude Code: teams, roles, autopilot and task routing across skills and subagents
README in short
The documentation shows how one line becomes a request, install steps, a command table, a list of automatic actions, supported models and links to symptom fixes and the FAQ. The main text is Korean, with English and Chinese versions.