How does this skill improve AI instruction following?

It audits your instructions for 'positive language' and 'strong constraints,' ensuring you use authoritative terms like 'MUST' and 'ALWAYS' which LLMs follow more reliably.

Can it detect broken links in my documentation?

Yes, it extracts file and command references and validates them against your actual filesystem and package.json to ensure the AI isn't being given hallucinated or outdated paths.

Does this work with tools other than Claude Code?

Yes, it is designed for cross-platform compatibility, supporting AGENTS.md which is used by OpenCode, Codex, and other AI development tools.

What files does this skill analyze?

It primarily targets CLAUDE.md and AGENTS.md files used by Claude Code, OpenCode, Codex, and other AI coding assistants to maintain project-specific context.

What is the 'lost in the middle' problem it addresses?

LLMs recall information best at the start and end of a file; this skill flags critical rules buried in the middle of long documents and suggests better positioning.

Claude Project Memory Enhancer

Name: Claude Project Memory Enhancer
Author: agent-sh

byagent-sh

•

692

•

生産性とワークフロー

Optimizes project memory files like CLAUDE.md and AGENTS.md to improve AI instruction following and context efficiency.

The enhance-claude-memory skill is a specialized diagnostic and optimization tool designed to audit the health of your project's AI memory files. By analyzing CLAUDE.md and AGENTS.md through the lens of advanced prompt engineering research, it identifies structural weaknesses, validates file and command references, and ensures critical instructions are positioned to avoid the 'lost in the middle' problem common in LLMs. This skill is essential for developers looking to maximize the reliability of AI assistants by enforcing positive constraints, strong language, and efficient token usage across their documentation.

主な機能

01Efficiency metrics including token count estimation and README duplication detection

02Content positioning analysis to ensure critical rules are placed in high-attention areas

03Automated validation of file paths and package.json commands referenced in memory files

04692 GitHub stars

05Instruction effectiveness auditing to convert negative constraints into high-compliance positive actions

06Cross-tool detection for CLAUDE.md and AGENTS.md files across various directory depths

ユースケース

01Standardizing AI instructions for cross-platform compatibility between Claude Code and OpenCode

02Auditing project memory files to improve Claude's adherence to specific coding standards

03Optimizing documentation to reduce token overhead while maintaining high instructional quality

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add agent-sh/agentsys enhance-claude-memory

For use in Claude.ai and ChatGPT

主な機能

01Efficiency metrics including token count estimation and README duplication detection

02Content positioning analysis to ensure critical rules are placed in high-attention areas

03Automated validation of file paths and package.json commands referenced in memory files

04692 GitHub stars

05Instruction effectiveness auditing to convert negative constraints into high-compliance positive actions

06Cross-tool detection for CLAUDE.md and AGENTS.md files across various directory depths

ユースケース

01Standardizing AI instructions for cross-platform compatibility between Claude Code and OpenCode

02Auditing project memory files to improve Claude's adherence to specific coding standards

03Optimizing documentation to reduce token overhead while maintaining high instructional quality