OpenClaw Hardening Baselines (Community)

Goal: give OpenClaw deployers a runnable baseline that reduces the most common enterprise failures:

• public / unsafe exposure
• over-privileged execution
• uncontrolled egress
• unreviewed skills/extensions

This repo is intentionally baseline-only (no platform, no SaaS).

What’s included

• Hardened Docker Compose starter (baselines/docker-compose/docker-compose.yml)
• Example reverse proxy config (TLS, auth boundary) (baselines/reverse-proxy/)
• Egress allowlist patterns + example firewall rules (docs/network-egress.md, policies/domain-allowlist.txt)
• Skill governance scaffolding (allowlist/blocklist templates) (policies/skill-*.yml, docs/skill-governance.md)
• act-check drift check that produces a simple JSON score (scripts/act-check.sh)

Quick start (Docker)

1. Copy the baseline: bash cp -r baselines/docker-compose ./openclaw-baseline cd openclaw-baseline

2. Review and set environment values in .env:

• bind gateway/control UI to 127.0.0.1 unless you really know why you need public access
• run behind a reverse proxy with strong auth if remote access is required

3. Start: bash docker compose up -d

4. Run drift check: bash ../scripts/act-check.sh --output ./act-check.json cat ./act-check.json

Feedback (help improve the baseline)

• If this baseline was useful, star the repo (it helps other teams find it).
• If you find gaps, open a GitHub Issue with:
  • your deployment context (Docker / VM / k8s)
  • which check failed (paste act-check.json with secrets redacted)
  • what you expected vs what happened
• PRs are welcome for additional hardening patterns and drift checks.

Support & warranty

No warranty. Use at your own risk. See LICENSE and docs/DISCLAIMER.md.

Contributing

PRs welcome. Please see CONTRIBUTING.md.