macrodata-distill

Extract actions and facts from OpenCode transcripts into a journal.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/ashwnn/agent-config --skill macrodata-distill-ashwnn
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: macrodata-distill
Source: https://github.com/ashwnn/agent-config/tree/main/skills/macrodata-distill
Command: npx skills add https://github.com/ashwnn/agent-config --skill macrodata-distill-ashwnn

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill consolidates and extracts actionable knowledge from daily OpenCode conversations by spawning dedicated sub-agents per conversation to prevent context blowup.

Core Features & Use Cases

  • Spawns sub-agents per conversation to distill actions, facts, and decisions.
  • Writes distilled results to a journal and updates entity files with new facts.
  • Handles only user and assistant content, filtering out tools and summaries.

Quick Start

Activate macrodata-distill to process today's conversations and generate distilled actions and facts.

Frequently Asked Questions about macrodata-distill

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

FAQPage Schema
How do I extract actionable knowledge from conversation transcripts?

To extract actionable knowledge from conversation transcripts, spawn a sub-agent per conversation to distill actions and facts, then write actions to a journal and aggregate facts by topic for entity updates. The coordinator pattern collects and filters these results.

How does the sub-agent pattern prevent context blowup when processing daily conversations?

The sub-agent pattern prevents context blowup by assigning each daily conversation to a dedicated sub-agent for distillation, allowing a coordinator to collect only the filtered actions and facts without overloading the context window.

What is the best way to filter non-actionable content from chat logs?

The best way to filter non-actionable content from chat logs is to process only user and assistant messages, excluding tools and summaries, while preserving verifiable technical details, file paths, and concrete outcomes.

Can I use macrodata-distill to update entity files with extracted facts?

Yes, you can use macrodata-distill to update entity files. It aggregates distilled facts per topic and writes them to entity files, while simultaneously routing extracted actions to a separate journal.

Does the conversation distillation process handle tool outputs and automated summaries?

No, the conversation distillation process does not handle tool outputs or automated summaries. It strictly filters those out, processing only user and assistant content to ensure safety and preserve verifiable details.