dk-compound

Extract conversation learnings into structured documents in docs/.

87|12|Updated Mar 1, 2026
One-click install
npx skills add https://github.com/deepklarity/harness-kit --skill dk-compound
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dk-compound
Source: https://github.com/deepklarity/harness-kit/tree/main/.claude/skills/dk-compound
Command: npx skills add https://github.com/deepklarity/harness-kit --skill dk-compound

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Captures learnings from conversations as a clean, human-readable document stored in docs/ to preserve decisions, patterns, and gotchas for future reuse.

Core Features & Use Cases

  • Structured learnings: extracts meta-level insights and concrete gotchas from a discussion.
  • Knowledge artifact generation: saves a reusable doc in docs/ that can be referenced in design reviews or post-mortems.
  • Activation workflow: supports a haiku subagent to analyze conversation and produce concise notes.

Quick Start

Direct an AI agent to capture the current conversation into a structured doc by running the activation command to generate the artifact in docs/.

Frequently Asked Questions about dk-compound

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

FAQPage Schema
How do I capture conversation learnings into reusable documentation?

Capture conversation learnings by running the activation command to trigger a haiku subagent, which analyzes the discussion, extracts meta-level principles and gotchas, and saves a structured human-readable document into the docs/ directory.

What is the best way to document design patterns discovered during a problem-solving session?

The best way to document design patterns is to run the activation command, which relies on a haiku subagent to analyze the conversation and produce concise, structured notes saved as a reusable document in the docs/ folder.

When should I generate a knowledge artifact from a conversation?

Generate a knowledge artifact after solving a non-trivial problem, discovering a design pattern, or encountering a gotcha that warrants lasting preservation. The Skill formalizes these conversation insights into a structured doc stored in docs/ for future reuse.

Does knowledge management documentation require any specific dependencies or setup?

No external dependencies are required for knowledge management documentation generation. It operates by directing an AI agent to run the activation command, which leverages a haiku subagent to analyze the conversation and output structured material into the docs/ directory.

What structure does the generated learning document follow?

The generated learning document enforces structured sections specifically for meta-level principles and concrete gotchas. It uses a haiku subagent to analyze the conversation and produce concise, clean, human-readable notes saved in the docs/ directory.