pkuppens/openclaw-security
skills/openclaw-security/SKILL.md
Enforces OpenClaw security constraints and mandates security audit before every code-change PR. Use when working on OpenClaw projects, before opening a PR, or when configuring or hardening OpenClaw.
development
Updated May 15, 2026
$ install --global
skillsauthnpx skillsauth add pkuppens/pkuppens openclaw-securityInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: May 15, 2026, 5:42 AM34.2s1 file scanned
SKILL.md
- name:
- openclaw-security
- description:
- Enforces OpenClaw security constraints and mandates security audit before every code-change PR. Use when working on OpenClaw projects, before opening a PR, or when configuring or hardening OpenClaw.
OpenClaw Security
Enforces the security posture required for OpenClaw. Mandatory for every code-change PR in OpenClaw projects.
When to use
- Before every PR that changes OpenClaw code or configuration
- When configuring or hardening an OpenClaw deployment
- When adding tools, gateways, or shared inboxes
- As part of quality-gate for OpenClaw repos
Mandatory PR requirement
Every code-change PR must include a passing OpenClaw security audit. Run before opening the PR:
openclaw security audit --deep
openclaw security audit --fix
Both must complete successfully. Do not bypass. If audit fails, fix issues before integration-commit and integration-pr.
Security constraints (non-negotiable)
OpenClaw is a hobby project in beta. Expect sharp edges. These constraints apply:
Trust boundary
- Personal agent by default — one trusted operator boundary.
- Bot can read files and run actions if tools are enabled.
- A bad prompt can trick it into doing unsafe things.
- OpenClaw is not a hostile multi-tenant boundary by default.
- If multiple users can message one tool-enabled agent, they share that delegated tool authority.
Before running
- If you are not comfortable with security hardening and access control, do not run OpenClaw.
- Ask someone experienced before enabling tools or exposing it to the internet.
Recommended baseline
| Control | Requirement |
|---------|-------------|
| Pairing/allowlists | Use pairing and mention gating |
| Multi-user/shared inbox | Split trust boundaries: separate gateway/credentials, ideally separate OS users/hosts |
| Sandbox | Use sandbox + least-privilege tools |
| Shared inboxes | Isolate DM sessions: session.dmScope: per-channel-peer; keep tool access minimal |
| Secrets | Keep secrets out of the agent's reachable filesystem |
| Model choice | Use the strongest available model for any bot with tools or untrusted inboxes |
Regular audit commands
Run regularly (at least before each PR):
openclaw security audit --deep
openclaw security audit --fix
Integration with existing skills
| Skill | Integration |
|-------|-------------|
| quality-gate | For OpenClaw repos: run openclaw security audit --deep and openclaw security audit --fix as mandatory checks before integration-commit. Report pass/fail in Quality Gate Results. |
| integration-pr | PR checklist must include: security audit passed. Do not open PR until audit passes. |
| integration-commit | Never commit without passing security audit when working on OpenClaw. |
Output format
Include in Quality Gate Results when applicable:
## OpenClaw Security Audit
- [x] openclaw security audit --deep — clean
- [x] openclaw security audit --fix — applied
If audit fails, list findings and remediation steps. Block commit/PR until resolved.
Reference
Must read: https://docs.openclaw.ai/gateway/security
Related Skills
pkuppens/sync-branch
testing
VerifiedTrustedCommunity
Syncs remote default branch locally (checkout, fetch --prune, pull) and returns to the previous branch when it still exists. Reports stashes and worktrees not yet handled. Use when the user asks to sync main, update default branch, fetch/pull origin, or run /sync-branch.
SKILL.mdUpdated Jun 6, 2026
pkuppens/azure-devops-work-items
tools
VerifiedTrustedCommunity
Creates, queries, updates, and links Azure Boards work items via az boards CLI. Use when filing ADO work items, running WIQL queries, or setting area path, iteration, tags, and assignee.
SKILL.mdUpdated May 29, 2026
pkuppens/azure-devops-repos
tools
VerifiedTrustedCommunity
Creates, reviews, and completes Azure Repos pull requests and branch policies via az repos CLI. Use when opening ADO PRs, setting required reviewers, or configuring build validation policies.
SKILL.mdUpdated May 29, 2026
pkuppens/azure-devops-pipelines
development
VerifiedTrustedCommunity
Guides Azure Pipelines YAML structure, build validation on PRs, and staged deployment with environments and approvals. Use when authoring azure-pipelines.yml or configuring CI/CD on Azure DevOps.
SKILL.mdUpdated May 29, 2026