skills
anthropics
Public repository for Agent Skills
Sahir619/fable-method
fable-method is a tool for DeepSeek Harness. The Fable Workflow: how Claude Fable 5 worked, distilled into skills any model can run, with the eval that keeps it honest. Think / act / prove.

REVIEW SCORECARD
✓ No risk detected · Automated scan results are for reference only, not an official endorsement
AI DEEP REVIEW
A well-evidenced, actively maintained agent-skill plugin with rigorous self-evaluation, worth installing for trap-heavy
README cites a CI checks badge, 15 eval rounds with 260+ agent runs, committed transcripts, and blind LLM judges that verify by diffing and executing rather than reading reports.
No download-and-execute or plaintext secrets are shown, and the method explicitly adds an authorization gate after an observed unauthorized staging deploy, though the install command pulls a marketplace plugin whose internals are not visible here.
Targets a real and common pain point (agents silently 'fixing' correct code, false completion claims, weak executors) with measured lift on traps, while honestly reporting null results on ordinary small tasks.
Last updated 2026-07-15, not archived, versioned at plugin v1.4.0, and backed by an ongoing eval log spanning 15 rounds.
README explains the think/act/prove/grow philosophy, the loop diagram, install command, and links to skill files and eval cases, but the excerpt is truncated mid-sentence and full usage guidance is only partially visible.
Generated by AI after reading the project README — a decision aid, not an endorsement. Neutral scores are given when information is limited.
Loading GitHub README...
The GitHub README is temporarily unavailable. Please try again later.
Open repositoryINSTALL REFERENCE
/plugin marketplace add Sahir619/fable-methodThis command was extracted from an explicit README install reference and is display-only; the DSH host will not execute it automatically.
PROJECT TOPICS
VALIDATION LADDER
FAQ
The Fable Workflow: how Claude Fable 5 worked, distilled into skills any model can run, with the eval that keeps it honest. Think / act / prove.
fable-method has 2,283 stars and 327 forks on GitHub, last updated 2026-07-15.
fable-method is distributed under the MIT license.
fable-method is listed in the DSH Universe directory as a tool for DeepSeek Harness. This plugin is listed in the DSH Universe directory and covered by its validation pipeline.
CLASSIFICATION EVIDENCE
The system checks GitHub Topics against the site taxonomy and keyword rules. Current matches:agent-skills。