⏳ This skill is pending AI review.
Scores will appear once the review pipeline completes.
code-navigate
Navigate code structure using grep-based heuristics — find definitions, references, call hierarchy, and symbol overviews. Use when tracing function calls, finding all usages of a symbol, understanding type hierarchies, or navigating unfamiliar codebases. Triggers on "go to definition", "find references", "where is this used", "what calls this", "trace this function", "understand this code".
Use with your AI agent
Open your project in any AI assistant that can read your files. Works with ChatGPT, Claude, Claude Code, Codex, Cursor, Hermes Agent, OpenClaw, Grok Bot, and more.
Your agent needs access to this page’s linked instructions and your project files. Copying does not install or execute anything.
// RATINGS
Not yet listed on ClawHub or SkillsMP
// README
OpenClaw Skills — Development Optimization Pack
A collection of skills that enhance OpenClaw for software development workflows. Inspired by capabilities found in modern AI coding assistants.
Skills
| Skill | Description |
|---|---|
| code-simplify | Three-way parallel code review: reuse, quality, and efficiency checks on your git diff |
| code-search | Fast codebase search using ripgrep + glob patterns |
| code-navigate | Grep-based code navigation — definitions, references, call hierarchy, symbol overview |
| memory-extract | Extract durable facts and decisions from conversations into persistent memory files |
| memory-organizer | Audit memory files for duplicates, outdated entries, and bloat |
| session-summary | Structured session progress capture — decisions, changes, open items |
| magic-docs | Self-maintaining documentation via # MAGIC DOC: headers |
Requirements
- OpenClaw installed and running
ripgrep(rg) recommended for code-search and code-navigate
# Debian/Ubuntu
apt-get install -y ripgrep
# macOS
brew install ripgrep
Installation
Install all skills
git clone https://github.com/EasyJoy-Technologies/openclaw-skills.git
cd openclaw-skills
./install.sh
Install a single skill
cp -r skills/<skill-name> ~/.openclaw/skills/<skill-name>
Verify
openclaw skills check
All installed skills should appear as 📦 (eligible).
Uninstall
# Remove a single skill
rm -rf ~/.openclaw/skills/<skill-name>
# Remove all skills from this pack
for d in skills/*/; do rm -rf ~/.openclaw/skills/$(basename "$d"); done
Recommended: Automatic Behaviors
For maximum value, add these rules to your agent's AGENTS.md (under ## Automatic Behaviors):
### Memory Extract (on task completion)
- **After completing a complex multi-step task**: write outcome, decisions, and new rules to memory.
- Do NOT extract trivial Q&A or information already in memory.
### Code Quality Review (on code changes)
- **After making code changes**: before committing, review the diff — check for duplicated logic, quality issues, and efficiency problems.
### Magic Docs (on relevant discovery)
- When reading a file that starts with `# MAGIC DOC:`, remember it.
- If new relevant information is discovered later, update the document.
### Code Search (prefer structured search)
- Prefer `rg` (ripgrep) over raw `grep`. Exclude `node_modules`, `.git`, `dist`, `build`.
How Skills Work
Skills are prompt-based modules. OpenClaw reads the SKILL.md frontmatter (name + description) to decide when to activate a skill. The body is loaded only when triggered.
- Skills live in
~/.openclaw/skills/<name>/ - Each skill has a
SKILL.mdwith YAML frontmatter - OpenClaw upgrades do not affect user skills (they're outside the install directory)
Contributing
- Fork this repo
- Add or improve a skill under
skills/<name>/SKILL.md - Follow the OpenClaw skill spec
- Submit a PR
License
MIT
// HOW IT'S BUILT
KEY FILES