Audit and improve an AI agent's skills, memory, and CLAUDE.md configurations.
npx skills add https://github.com/kiloloop/oacp-skills --skill self-improve-kiloloop
Please help me install this Agent Skill. Skill: self-improve Source: https://github.com/kiloloop/oacp-skills/tree/main/skills/self-improve/claude Command: npx skills add https://github.com/kiloloop/oacp-skills --skill self-improve-kiloloop
This Skill proactively identifies and fixes issues within an AI agent's core components, ensuring its skills, memory, and configurations remain accurate, up-to-date, and efficient.
/self-improve identifies these issues, suggests corrections, and updates the agent's knowledge base, restoring optimal performance.Run a full review of the agent's skills, memory, and configuration files.
High-intent search queries and answers about installing and using this skill.
To audit AI agent skills and memory files, you can run a full review to detect staleness, contradictions, and bloat across configurations. The system identifies issues within skills, memory files, and CLAUDE.md configurations, then proposes and applies fixes to restore optimal performance.
AI agent performance degradation is often caused by outdated skill instructions or conflicting memory entries. You can resolve this by running a targeted review to identify these specific issues, suggest corrections, and commit updates to the agent's knowledge base repository.
Yes, you can perform targeted reviews of specific skills or file types instead of a full audit. This allows you to isolate and validate individual components like memory files or CLAUDE.md configurations, checking for missing safety rules or duplication before applying fixes.
Memory hygiene for AI operations is the process of detecting stale, conflicting, or inconsistent data across agent memory files. You need it when an agent's knowledge base accumulates bloat or contradictions, ensuring configurations remain accurate and efficient through automated cleanup.
To remove configuration bloat and duplications from CLAUDE.md files, run a configuration validation scan. This process identifies redundant settings and contradictions, then proposes and applies approved changes, committing the streamlined configurations to the correct repository.
az9713Initialize a project with dark code prevention directories, CLAUDE.md section, and baseline audit.
AvatarGanymedeSynchronize .kandev/ configuration files to the office database.
cajiasMonitor GitHub pull requests for new commits and resume review workflows.
xfs96192Extract patterns from skill interactions and update related skills with versioned templates.