Contribute to the OpenClaw core repository: pick issues, inspect bundled skills, make focused changes, run the right level of validation, and open a clean PR. Use when working specifically on openclaw/openclaw or a fork. Not for generic git usage or unrelated GitHub repos.
Installs into .claude/skills of the current project.
Are you the author of Openclaw Contributor?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/arry8-openclaw-contributor)
---
name: openclaw-contributor
description: "Contribute to the OpenClaw core repository: pick issues, inspect bundled skills, make focused changes, run the right level of validation, and open a clean PR. Use when working specifically on openclaw/openclaw or a fork. Not for generic git usage or unrelated GitHub repos."
metadata:
{
"openclaw":
{
"emoji": "π¦",
"requires": { "bins": ["gh", "git", "pnpm"] },
"install":
[
{
"id": "brew-gh",
"kind": "brew",
"formula": "gh",
"bins": ["gh"],
"label": "Install GitHub CLI (brew)",
},
],
},
}
---
# OpenClaw Contributor
Use this skill when contributing directly to `openclaw/openclaw` or a fork of it.
## When to Use
β **USE this skill when:**
- fixing an OpenClaw bug or implementing an OpenClaw feature
- adding or improving a bundled skill under `skills/`
- preparing a clean fork β branch β commit β PR workflow
- checking nearby code, docs, tests, and issue context before changing anything
- coordinating multiple agents for research, implementation, and PR polish
## When NOT to Use
β **DON'T use this skill when:**
- doing generic local git work unrelated to OpenClaw
- working on a different repository with different contribution norms
- using GitHub only for issues/PR status β use the `github` skill
- doing broad autonomous issue farming across many repos β use `gh-issues`
## Quick Orientation
Important paths in the OpenClaw repo:
```bash
skills/ # bundled skills
packages/ # package sources (if present in current layout)
docs/ # documentation
scripts/ # build, release, and utility scripts
package.json # scripts: build, lint, test, dev
```
Useful first checks:
```bash
# Verify auth
gh auth status
# See open issues
gh issue list --repo openclaw/openclaw --state open --limit 30
# Inspect a bundled skill
ls skills/
sed -n '1,160p' skills/github/SKILL.md
# See repo scripts
node -e "const p=require('./package.json'); console.log(p.packageManager); console.log(p.scripts)"
```
## Standard Contribution Flow
### 1. Start from a fork
```bash
gh repo fork openclaw/openclaw --clone
cd openclaw
git remote add upstream https://github.com/openclaw/openclaw.git 2>/dev/null || true
git fetch upstream
```
### 2. Create a focused branch
```bash
git checkout -b feat/my-change
# or
# git checkout -b fix/my-change
# git checkout -b docs/my-change
```
### 3. Inspect before editing
Before changing anything:
- read the nearest comparable file first
- preserve the existing repo tone and structure
- keep skill directories minimal: normally just `SKILL.md`, plus `scripts/` or `references/` only when truly needed
- do **not** add extra files like `README.md`, `CHANGELOG.md`, or installation guides inside a skill directory
For skill work specifically:
```bash
sed -n '1,200p' skills/skill-creator/SKILL.md
sed -n '1,200p' skills/github/SKILL.md
```
### 4. Make the smallest useful change
Guidelines:
- prefer one clear improvement over a grab-bag PR
- match existing command style and frontmatter format
- avoid hype, speculation, or unverifiable claims in bundled content
- if adding a skill, make the description triggerable and specific
- if instructions vary by scenario, keep the overview in `SKILL.md` and move heavy detail to `references/`
### 5. Run the right level of validation
For **docs / markdown / skill-only** changes:
```bash
git diff --check
```
For **runtime / CLI / core** changes:
```bash
pnpm lint
pnpm test
```
For **targeted investigation** before full test runs:
```bash
node -e "const p=require('./package.json'); console.log(p.scripts.test)"
```
Do not claim tests were run if they were not.
### 6. Commit clearly
```bash
git add <files>
git commit -m "feat(skills): add openclaw contributor skill"
```
Good commit / PR title patterns:
- `feat(skills): add openclaw contributor skill`
- `fix(gateway): reduce repeated file read context bloat`
- `docs(skills): clarify bundled skill structure`
### 7. Push and open PR
```bash
git push -u origin HEAD
gh pr create \
--repo openclaw/openclaw \
--title "feat(skills): add openclaw contributor skill" \
--body-file .github/pr-body.md
```
If no PR body file exists, write one first.
## PR Body Template
Create a short, factual body:
```markdown
## Summary
- add a bundled `openclaw-contributor` skill
- document repo-specific contribution workflow for OpenClaw contributors
## Why
OpenClaw ships many bundled skills but does not yet include one focused on contributing back to the OpenClaw repo itself.
## Validation
- reviewed nearby bundled skills for structure and tone
- ran `git diff --check`
```
Only mention `pnpm lint` / `pnpm test` if you actually ran them.
## Multi-Agent Pattern
For larger changes, split work into three roles:
1. **Researcher** β issue triage, codebase reconnaissance, similar-file scan
2. **Implementer** β smallest correct patch, local validation
3. **Reviewer** β PR body, risk scan, regression checklist
Keep one agent responsible for final judgment and commit quality.
## Handy Commands
```bash
# Find candidate issues
gh issue list --repo openclaw/openclaw --state open --limit 50
# View one issue
gh issue view 12345 --repo openclaw/openclaw
# See changed files before commit
git diff --stat
# Check branch state
git status --short
# See recent upstream changes
git log --oneline --decorate --graph -20
```
## Notes
- Bundled skills should feel production-ready, not personal.
- Prefer repo-specific accuracy over generic open-source advice.
- If a contribution depends on undocumented assumptions, open an issue instead of guessing.
- When in doubt, ship the smaller PR first.