Does the export include performance metrics?

Yes, it automatically attaches confidence scores, autonomy levels, and total session counts to each exported playbook to provide context on the AI's proficiency.

What is the ALM Export skill?

It is a utility within the Claude Code Learning Memory System that allows users to export learned playbooks into a portable, standalone Markdown format for external use.

Is the exported format compatible with Git?

Absolutely. The clean Markdown output is designed specifically to be committed to version control systems like Git to track how your AI's logic evolves over time.

Where does the skill source the playbook data from?

The skill reads .md files and metadata from the standard Claude configuration directory, typically located at ~/.claude/alm/playbooks/.

Can I export specific playbooks instead of everything?

Yes, you can specify a specific task type as an argument to export a single playbook, or use the 'all' argument to export your entire library.

ALM Playbook Exporter

Name: ALM Playbook Exporter
Author: sfw

bysfw

0•

学习与文档

Exports Claude Code Learning Memory System (ALM) playbooks into clean, shareable Markdown files for documentation and version control.

ALM Playbook Exporter is a specialized utility designed to bridge the gap between automated AI learning and human-readable documentation. It retrieves localized playbooks from the Claude Code Learning Memory System (ALM), attaches critical metadata such as confidence scores and session history, and formats them into standardized Markdown. This skill is essential for developers who need to version control their AI's learned behaviors, share organizational best practices with team members, or manually refine the logic Claude has developed through iterative task performance.

主要功能

01Exports playbooks as standalone, human-readable Markdown files

020 GitHub stars

03Facilitates version control and manual editing of learned AI workflows

04Includes confidence scores, autonomy levels, and session statistics in the header

05Supports selective export by specific task type or bulk export of all playbooks

06Integrates global playbook data and metadata from confidence records automatically

使用场景

01Auditing and manually refining AI-generated playbooks to ensure architectural compliance

02Version controlling Claude's evolving strategies in a Git repository for audit trails

03Sharing learned project workflows and best practices with remote teammates

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add sfw/alm export

For use in Claude.ai and ChatGPT

Download Skill