agent-ops-optimize-instructions
About
This skill optimizes agent instruction files by extracting verbose sections into separate reference files and replacing them with concise summaries. It reduces overall context size to stay within token limits while preserving all original information. Use it when instruction files grow too large, typically exceeding ~2000 lines or containing extensive code examples.
Quick Install
Claude Code
Recommendednpx skills add majiayu000/claude-skill-registry -a claude-code/plugin add https://github.com/majiayu000/claude-skill-registrygit clone https://github.com/majiayu000/claude-skill-registry.git ~/.claude/skills/agent-ops-optimize-instructionsCopy and paste this command in Claude Code to install this skill
GitHub Repository
Related Skills
agent-ops-migrate
OtherThis skill migrates projects between repositories while ensuring functionality and validating complete content transfer. It's designed for monorepo consolidation, template upgrades, and codebase mergers, handling conflicts and providing rollback capabilities. The workflow uses multiple agent-ops skills to manage project sections, establish baselines, validate transfers, and resolve issues.
agent-ops-github
OtherThis skill enables bidirectional synchronization between agent-ops issues and GitHub Issues, allowing developers to push local issues to GitHub or pull remote issues locally. It's useful for sharing work with teams, importing PR feedback as issues, and keeping both systems in sync. The skill works with or without the `aoc` CLI and can monitor external repositories.
agent-ops-research
OtherThis skill conducts structured, in-depth research on technical topics like technologies, libraries, or patterns, producing documented findings. It can optionally create issues from its research to track actionable items. Use it when evaluating new tools, investigating solutions, or preparing for design decisions.
agent-ops-idea
OtherThis skill transforms raw, unstructured ideas into researched and documented backlog issues. It's triggered when users present vague concepts that need exploration and fleshing out. The process involves capturing the initial idea, enriching it with research, and generating a formal "IDEA" issue for future triage.
