Does it filter out noise from Pull Requests?

Yes, it is designed to ignore 'LGTM' comments, bot messages, and simple questions, focusing instead on improvement suggestions, pattern mentions, and bug identifications.

What does the /compound command do?

It triggers a workflow that scans your current Pull Request and local development specs to extract lessons learned, saving them as structured markdown files in your documentation folder.

Where are the resulting documents saved?

Documents are saved to the docs/learnings/ directory, using a standardized naming convention of YYYY-MM-DD-title.md for easy chronological browsing.

Which sources does this skill use for data collection?

It pulls information from GitHub PR comments, review feedback, and specific local files like learnings.md, decisions.md, and issues.md located in your spec directories.

Knowledge Compounder

Name: Knowledge Compounder
Author: team-attention

byteam-attention

•

Learning & Documentation

Extracts and structures technical learnings and architectural decisions from pull requests and development context into persistent documentation.

Knowledge Compounder streamlines the post-development documentation process by gathering insights from GitHub PR comments, peer reviews, and internal context files. It intelligently synthesizes this data into standardized markdown documents within your repository, ensuring that valuable engineering lessons, design patterns, and bug-fix rationales are preserved as searchable organizational memory rather than being lost in closed pull requests.

Key Features

01Supports cross-referencing between related documentation for better discovery

02Extracts high-value feedback and suggestions from GitHub PR comments

0364 GitHub stars

04Classifies technical learnings by problem type and generates metadata tags

05Automates document generation using standardized markdown templates

06Aggregates data from local spec files including decisions.md and issues.md

Use Cases

01Building an automated repository of engineering best practices for team onboarding

02Documenting recurring bug patterns and their solutions identified during code review

03Capturing architectural trade-offs and rationale after completing a major feature

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add team-attention/hoyeon compound

For use in Claude.ai and ChatGPT

Download Skill