skills.rest
leadmind-ai
leadmind-aiOFFICIAL
Explore all skills by this author →
GitHub

feedback-loop

Extract recurring friction patterns from logs and generate rules for CLAUDE.md.

Updated Mar 25, 2026
Search Tags:#process-improvement#git-history#rule-generation#friction-analysis#production-rules#qa-reports
One-click install
npx skills add https://github.com/leadmind-ai/leadmind-ai-website --skill feedback-loop-leadmind-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: feedback-loop
Source: https://github.com/leadmind-ai/leadmind-ai-website/tree/main/.claude/skills/feedback-loop
Command: npx skills add https://github.com/leadmind-ai/leadmind-ai-website --skill feedback-loop-leadmind-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill analyzes frictions encountered during development (friction logs, git history, QA/security reports, and code TODO/FIXME notes) and turns them into concrete, actionable rules for CLAUDE.md or .claude/rules/. It enables continuous improvement loops that help teams preempt recurring issues.

Core Features & Use Cases

  • Collects friction signals from multiple sources (docs/friction-log.md, git history, security and QA reports, code comments) and identifies repeating patterns.
  • Converts patterns into explicit rules that guide production behavior and team coordination.
  • Supports deployment of rules to CLAUDE.md under "Règles de production" or to .claude/rules/ for team-wide adoption, enabling faster remediation in future slices.

Quick Start

Provide a friction log or run a scan to generate the first set of actionable rules and start tightening the development loop.

Frequently Asked Questions about feedback-loop

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I convert development friction logs into actionable rules for CLAUDE.md?▼

To convert friction logs into actionable rules, this Skill analyzes friction logs, git history, QA reports, and code comments to identify recurring patterns and generates explicit rules for deployment to CLAUDE.md or .claude/rules/.

What is the best way to extract recurring friction patterns from git history and code comments?▼

The best way to extract recurring friction patterns is scanning multiple sources like git history, security reports, and TODO/FIXME notes to identify repeating issues and transform them into concrete production rules for team coordination.

Can I use QA and security reports to generate production rules for .claude/rules/?▼

Yes, you can use QA and security reports to generate production rules. The Skill collects friction signals from these reports alongside friction logs and git history to create traceable rules for .claude/rules/.

Do I need a friction log to start generating rules from code comments and git history?▼

You do not strictly need a friction log to start generating rules. You can run a scan across git history, QA reports, and code TODO/FIXME comments to extract patterns and generate your first set of actionable rules.

How does transforming project frictions into rules help continuous process improvement?▼

Transforming project frictions into rules helps continuous process improvement by creating a feedback loop that preempts recurring issues, enabling faster remediation in future development slices through team-wide rule adoption.

Related Skills

Explore More Software Engineering Skills→
s0613s0613
1

sj-retro

Aggregate git, test, and friction data into structured engineering retro reports.

Community
Intermediate
jarrett-aujarrett-au

codebase-rules-generator

Analyze codebase architecture and patterns to generate Claude Code rules and CLAUDE.md files.

Community
Intermediate
Arnutt-NArnutt-N

security-scanner

Detect OWASP Top 10 risks and misconfigurations in HR-IMS code.

Community
Advanced
IntegralistIntegralist
2

durable-rules

Analyze investigation results to codify systemic patterns as conventions or anti-patterns.

Community
Advanced