history-hygiene

Record final decision outcomes to history.md after confirmation.

Updated Nov 11, 2025
One-click install
npx skills add https://github.com/acmeleme/Conversation-Knowledge-Mining-Solution-Accelerator --skill history-hygiene-acmeleme
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: history-hygiene
Source: https://github.com/acmeleme/Conversation-Knowledge-Mining-Solution-Accelerator/tree/main/.squad/templates/skills/history-hygiene
Command: npx skills add https://github.com/acmeleme/Conversation-Knowledge-Mining-Solution-Accelerator --skill history-hygiene-acmeleme

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

History files (.md files tracking decisions, spawns, outcomes) are read cold by future agents. Stale or incorrect entries poison decision-making downstream. The Kobayashi incident proved this: history said "Brady decided v0.6.0" when Brady had reversed that to v0.8.17. Future spawns read the wrong truth and repeated the mistake.

Core Features & Use Cases

  • Record the final outcome to history.md, not intermediate states.
  • Wait for confirmation before writing to history — avoid premature entries.
  • Update history immediately if a decision reverses, ensuring a single source of truth for future tasks.

Quick Start

Finalize a decision and log the final outcome to history.md with the date and a concise summary.

Frequently Asked Questions about history-hygiene

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

FAQPage Schema
How do I record final decisions in a history.md file?

To record final decisions in a history.md file, log the final outcomes with a date and concise summary only after confirmation, ensuring future agents read a single truth.

Why should I avoid logging intermediate AI requests in history files?

Avoiding intermediate AI requests in history files prevents stale or incorrect entries from poisoning downstream decision-making, ensuring future spawns read only confirmed and finalized outcomes.

What is the best way to handle reversed decisions in markdown history logs?

The best way to handle reversed decisions in markdown history logs is to update history.md immediately upon reversal, replacing stale entries to maintain a single source of truth for future tasks.

When do I need to update traceability logs for AI agents?

You need to update traceability logs for AI agents immediately after a decision is confirmed or reversed, preventing future spawns from reading incorrect truth and repeating past mistakes.

Does writing to history.md require waiting for confirmation?

Writing to history.md requires waiting for confirmation to avoid premature entries, applying this across decision workflows to guarantee future agents read only finalized outcomes.

What happens if history.md contains stale decision entries?

If history.md contains stale decision entries, future agents read the wrong truth and repeat mistakes during downstream decision-making, which is why logging only final outcomes is enforced.