managing-signs

Extract learnings from session transcripts and apply approved signs to CLAUDE.md files.

Updated Jan 14, 2026
One-click install
npx skills add https://github.com/Mission42-ai/m42-claude-plugins --skill managing-signs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: managing-signs
Source: https://github.com/Mission42-ai/m42-claude-plugins/tree/main/plugins/m42-signs/skills/managing-signs
Command: npx skills add https://github.com/Mission42-ai/m42-claude-plugins --skill managing-signs

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This skill provides a structured workflow to extract learnings from session transcripts and apply them as signs in CLAUDE.md files, helping teams capture and reuse insights.

Core Features & Use Cases

  • Extract learnings from session transcripts and populate a backlog for review
  • Review and approve learnings, then apply them to CLAUDE.md files
  • Maintain a standardized backlog schema and sign format for consistency across projects
  • Integrate with sprint workflows to automate knowledge capture

Quick Start

Install/use the skill in your Claude Code environment and begin by running the extraction on a sample transcript:

  • Run /m42-signs:extract path/to/transcript.jsonl --confidence-min medium
  • Run /m42-signs:review to approve learnings
  • Run /m42-signs:apply to apply approved signs to CLAUDE.md files

Frequently Asked Questions about managing-signs

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

FAQPage Schema
How do I extract learnings from session transcripts and add them to CLAUDE.md?

You can extract learnings from session transcripts by running an automated extraction command on the transcript file, then reviewing the generated backlog and applying approved items as structured signs directly into CLAUDE.md files.

What is a learning backlog and how does it help manage CLAUDE.md signs?

A learning backlog is a structured schema that captures extracted insights from session transcripts for review. It enforces consistent sign templates to ensure traceable improvements before applying them to CLAUDE.md files.

Can I integrate transcript extraction and sign reviews into sprint workflows?

Yes, the extraction and review process supports sprint workflows to automate knowledge capture. You can run extraction on transcripts, review the generated backlog, and apply approved signs to keep CLAUDE.md files updated during sprints.

How do I review and approve extracted learnings before applying them?

After extracting learnings from a transcript into the backlog, you run a review command to evaluate them. You can set a minimum confidence level during extraction, then approve learnings to apply them as signs to CLAUDE.md files.

Does this approach require a specific format for session transcripts and signs?

It processes session transcripts in JSONL format and enforces a standardized backlog schema alongside sign templates. This structured format ensures consistent, traceable improvements when applying signs to CLAUDE.md files.

What's the best way to turn session insights into actionable knowledge for Claude?

The best way is to automate extraction from session transcripts into a standardized backlog, review and approve the captured learnings, and apply them as structured signs in CLAUDE.md files for consistent reuse.