What triggers the documentation process?

The skill triggers automatically when phrases like 'that worked' or 'problem solved' are detected in the chat, or it can be manually invoked via the /doc-fix command.

What is the benefit of promoting a doc to 'Critical Patterns'?

Promoting a solution to Critical Patterns ensures it is added to a high-priority 'Required Reading' file that AI subagents scan before generating code, preventing the repetition of known mistakes.

Does it support metadata validation?

Yes, it includes a blocking YAML validation step that ensures all metadata matches a predefined schema before the documentation file is finalized.

Where are the documentation files stored?

Solutions are saved as markdown files within categorized subdirectories under the docs/solutions/ path in your project repository.

Can I link related issues together?

Absolutely. During the process, the skill searches for similar issues and provides a menu to cross-reference new docs with existing ones or update them if the root cause is the same.

Compound Docs

Name: Compound Docs
Author: gvkhosla

bygvkhosla

•

Learning & Documentation

Captures and categorizes solved technical problems into structured, searchable markdown documentation with validated YAML metadata.

Compound Docs streamlines the creation of a persistent institutional knowledge base by automatically documenting technical resolutions immediately after a fix is confirmed. It enforces a strict 7-step capture process—including automated context gathering, duplicate checking, and YAML schema validation—to ensure every entry is high-quality and machine-readable. By organizing solutions into categorized directories and offering promotion to 'Critical Patterns,' it transforms transient chat history into a durable resource that improves the accuracy of future AI sessions and developer workflows.

Key Features

01Post-capture decision menu for promoting solutions to 'Critical Patterns'

02Context extraction for symptoms, root causes, and prevention strategies

03Automatic invocation triggered by common problem-resolution phrases

04Strict YAML frontmatter validation against schema enums for consistent metadata

05Automated directory organization based on specific problem types

0635 GitHub stars

Use Cases

01Building a searchable repository of complex debugging solutions for specialized tech stacks

02Preventing recurring bugs by documenting non-obvious root causes and required fixes

03Onboarding AI agents with 'Required Reading' lists derived from verified project resolutions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gvkhosla/compound-engineering-pi compound-docs

For use in Claude.ai and ChatGPT

Download Skill