Miguok GitHub avatar

fable-harness

Miguok

Fable Harness is a drop-in behavior protocol that injects a disciplined engineering process—gather evidence, state assumptions, challenge conclusions, and prove with tests—into every Claude Code session.

Stars

197

7-day growth

No data

Forks

35

Open issues

0

License

MIT

Last updated

2026-07-20

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It distills the careful procedure from Anthropic's Fable model into a reusable kit, addressing the gap between good and sloppy agent behavior by enforcing consistent procedural discipline across Claude Opus, Sonnet, and Haiku.

Who it is for

  • Developers using Claude Code for software engineering
  • Teams seeking reliable AI-assisted development workflows
  • Engineers wanting to reduce mistakes from hasty agent decisions
  • Project leads standardizing agent behavior across a team

Use cases

  • Enforce evidence-gathering before answering coding questions
  • Run adversarial review on architecture decisions before committing
  • Ensure automated tests are written and pass for every code change
  • Route reasoning tasks to expensive models and bulk edits to cheaper ones

Strengths

  • OODA loop forces Claude to read real files instead of guessing from memory
  • Three-opponent adversarial review (skeptic, red-team, simplifier) reduces flawed conclusions
  • Tiered model routing optimizes cost by matching task complexity to model capability
  • Token efficiency via isolated sub-agents and parallel execution without benchmarked figures

Considerations

  • Transplants only procedure, not the model's innate judgment
  • Cost savings described as architectural reasons, not benchmarked percentages
  • Requires Claude Code environment and may conflict with existing harnesses like harnessmith or Superpowers

README quick start

Quick start

Description

Make Claude Code work like a disciplined engineer: OODA, multi-party adversarial review, model routing, fail-then-pass. Distilled from Fable to reinforce the Opus harness.

Related repositories

Similar projects matched by category, topics, and programming language.

lopopolo
Featured
lopopolo GitHub avatar

harness-engineering

Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.

AI & Machine LearningAI Agents
2,390
slvDev
Featured
slvDev GitHub avatar

esp32-ai

A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.

AI & Machine LearningLarge Language Models
1,960
littledivy
Featured
littledivy GitHub avatar

mimic

mimic captures traffic from any iOS or web app and automatically generates a Python client library that lets you call the app's API like a regular library.

AI & Machine Learning
1,482