Self Review

Audit weekly agent outputs, reliability, and memory hygiene into a markdown report.

Updated Jun 2, 2026

Search Tags:#automation #logs #audit #memory #reporting #improvement #self-review

Ready to run

npx skills add https://github.com/Atrium-Hermes/atrium-lighthouse --skill self-review-atrium-hermes

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: Self Review
Source: https://github.com/Atrium-Hermes/atrium-lighthouse/tree/main/skills/self-review
Command: npx skills add https://github.com/Atrium-Hermes/atrium-lighthouse --skill self-review-atrium-hermes

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Weekly automated review of agent performance, highlighting what worked, what failed, and what to improve.

Core Features & Use Cases

Audit outputs for quality and identify patterns of errors or success.
Evaluate reliability by tracking skill execution and failure signals.
Maintain memory hygiene by checking memory and logs for freshness and structure.

Quick Start

Run the self-review workflow to audit outputs, reliability, and memory hygiene, generate improvements, and save the report.

Frequently Asked Questions about Self Review

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

FAQPage Schema

How do I automate a weekly audit of agent activity logs and memory?▼

Weekly agent activity audits are automated by analyzing logs, memory, and PR history from the last 7 days to identify quality issues and generate improvement recommendations. The workflow outputs a full review report and logs the results to memory.

What is memory hygiene and how does an automated self-review maintain it?▼

Memory hygiene is the practice of checking memory and logs for freshness and structure. The self-review evaluates memory hygiene, prunes memory as needed, and logs the outcome to ensure agent memory remains clean and reliable.

How do I track agent reliability and identify patterns of errors or success?▼

Agent reliability is tracked by evaluating skill execution and failure signals from the last 7 days of logs. The audit identifies patterns of errors and success to highlight what worked, what failed, and what to improve.

Can I save automated agent performance reviews to a markdown file and send a summary notification?▼

Automated agent performance reviews are saved to articles/self-review-${today}.md, and a summary is sent via the notify command. The process also logs results to memory/logs/${today}.md for record-keeping.

Does the weekly self-audit require any external dependencies or components to run?▼

The weekly self-audit requires no external dependencies or components to run. It operates independently to analyze memory, logs, article sources, and PR history to generate improvement recommendations.