learn

Analyze user feedback and propose scoped updates to workspace context documentation.

16|2|Updated Jun 17, 2026
One-click install
npx skills add https://github.com/pipecrew-ai/pipecrew --skill learn-pipecrew-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: learn
Source: https://github.com/pipecrew-ai/pipecrew/tree/main/skills/learn
Command: npx skills add https://github.com/pipecrew-ai/pipecrew --skill learn-pipecrew-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the challenge of capturing and applying user feedback to update workspace context documentation, facilitating knowledge retention and process improvement over time.

Core Features & Use Cases

  • Feedback Analysis: Analyze merged PRs, recorded deliveries, branch diffs, Claude Code sessions, and free-form text to identify patterns and areas for improvement.
  • Scoped Updates: Propose updates to durable context documents like platform.md, CLAUDE.md, and DESIGN_SYSTEM.md based on the analysis.
  • Tier Classification: Classify findings into workspace, repo, plugin, and run-local scopes for targeted updates.
  • Fix-Round Dispatch: Optionally dispatch implementer agents to apply findings to affected branches for immediate alignment with new conventions.

Quick Start

Run the /pipecrew:learn command with the appropriate flags, such as --pr=<url> or --run=<run_id>, to initiate the learning process.

Frequently Asked Questions about learn

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

FAQPage Schema
How do I update workspace context docs from user feedback?

To update workspace context docs from user feedback, analyze merged PRs and session logs to propose scoped updates to files like platform.md and CLAUDE.md. This process classifies findings into workspace and repo tiers to ensure knowledge retention.

What is the best way to retain engineering knowledge across multi-repo workflows?

Retaining engineering knowledge across multi-repo workflows involves analyzing branch diffs and recorded deliveries to identify process improvements. The system proposes targeted updates to durable context documents, ensuring critical conventions persist across repositories.

Can I automatically apply feedback analysis findings to affected branches?

Yes, you can automatically apply feedback analysis findings to affected branches. The fix-round dispatch feature optionally deploys implementer agents to update branches directly, aligning them immediately with newly established conventions.

How does feedback analysis classify documentation updates?

Feedback analysis classifies documentation updates by sorting findings into four distinct scopes: workspace, repo, plugin, and run-local. This tier classification ensures that proposed changes to context documents target the correct architectural level.

Do I need Python scripts to analyze Claude Code sessions for process improvement?

Yes, you need Python scripts to analyze Claude Code sessions and free-form text for process improvement. These scripts perform the feedback analysis required to identify patterns and propose scoped updates to your context documentation.