Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
This skill manages an immutable-config workflow where changes to local configuration files do not take effect until they are explicitly released. It provides a structured process for updating settings and fixing configuration bugs to ensure changes reach production.
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Follow the documented setup, then try a first task.
I’m using
Claude Code plugin
Installs: Complete Fable Workflow plugin, containing all four namespaced skills: fable-method, fable-loop, fable-judge, and fable-domain. · Claude Code
Before you start
Claude Code with plugin support
Enter both commands, in order, inside a Claude Code session.
The failing test name or path and its failure output
Relevant specification or documentation and the project's test command
Use fable-method in plan mode for the failing test I identify. Read the test, the implementation it exercises, and the relevant specification. Give me a cited diagnosis, a definition of done with its verification, and one recommended approach. Stop after the plan without editing files.
The local repository and diff, or a changed directory with a pristine reference
The environment and commands needed to reproduce the claimed local checks
Use fable-judge in default mode to verify the completed fix and report I provide. Compare the actual diff with the original request, re-run the claimed local checks, and inspect changed tests for weakened assertions. Return the verdict first, followed by a claims-and-observations table and the smallest recommended fix if needed. Leave files unchanged and identify any verification you cannot reproduce.
A local project with one reproducible failing test
The failure output, intended behavior, and relevant specification
A working local test environment and the relevant test command
An agent environment supporting subagents, with fable-method installed alongside fable-loop
Use fable-loop to fix the single failing test I identify in this local project. Gather evidence from the test, implementation, and specification before deciding what should change; surface any contradiction. Make the smallest correct reversible change, run the relevant checks, and report the observed outcome with caveats. Keep the scope to this failure and local verification.
Claude
Codex
Cursor