prompt-regression

Community

Detect harness behavior regressions before they ship.

Authorchrono-meta
Version1.0.0
Installs0

System Documentation

What problem does it solve?

When you edit core harness rules, update skills, or modify trigger mappings, AI behavior can silently break without warning, leading to unexpected responses and broken workflows in your Claude Code projects. This skill eliminates that risk by automatically validating behavioral correctness after every change.

Core Features & Use Cases

  • Targeted Probe Testing: Runs only the regression probes relevant to your changed files, avoiding unnecessary full-suite runs for small edits.
  • Baseline Comparison: Compares current AI response patterns against saved baselines to spot even subtle behavioral shifts.
  • Use Case: After you update a skill's trigger phrases or modify CLAUDE.md core onboarding rules, run this skill to confirm the AI still initiates the correct workflows and follows required chains when given standard prompts.

Quick Start

Use the prompt-regression skill to verify no behavioral regressions were introduced after you make changes to your harness rules or skills.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: prompt-regression
Download link: https://github.com/chrono-meta/forge-harness/archive/main.zip#prompt-regression

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 536,000+ vetted skills library on demand.