security-sweep

# Security Sweep — Skill & Plugin Auditor Scans OpenClaw skills and plugins for: 1. **Hardcoded secrets** — API keys, tokens, passwords in code 2. **Dangerous exec patterns** — shell injection, eval, unsanitized child_process calls 3. **Dependency vulnerabilities** — npm audit failures 4. **Network egress** — unexpected outbound connections 5. **Input injection** — unsanitized user input reaching exec/file/eval ## Scan Scope **Built-in skills** (read-only, bundled with OpenClaw CLI): ``` $(brew --prefix)/Cellar/openclaw-cli/<version>/libexec/lib/node_modules/openclaw/skills/ ``` **Workspace skills** (user-installed): ``` ~/.openclaw/workspace/skills/ ``` ## Workflow ### Full Sweep Run the comprehensive scan script: ```bash SKILLS_DIR="$(brew --prefix)/Cellar/openclaw-cli/2026.3.24/libexec/lib/node_modules/openclaw/skills" WS_DIR="$HOME/.openclaw/workspace/skills" REPORT_DATE=$(date +%Y%m%d_%H%M%S) REPORT_FILE="$HOME/.openclaw/security-sweep-${REPORT_DATE}.txt" bash ~/.openclaw/workspace/skills/security-sweep/scripts/full-scan.sh \ --builtin "$SKILLS_DIR" \ --workspace "$WS_DIR" \ --output "$REPORT_FILE" ``` ### Quick Scan (fast patterns only) ```bash bash ~/.openclaw/workspace/skills/security-sweep/scripts/quick-scan.sh \ --dir "$HOME/.openclaw/workspace/skills" ``` ### Single Skill Scan ```bash bash ~/.openclaw/workspace/skills/security-sweep/scripts/skill-scan.sh \ --skill /path/to/skill ``` ### NPM Audit (workspace skills with package.json) ```bash bash ~/.openclaw/workspace/skills/security-sweep/scripts/npm-audit.sh \ --workspace "$HOME/.openclaw/workspace/skills" ``` ## Risk Categories | Level | Finding | Action | |-------|---------|--------| | 🔴 CRITICAL | Hardcoded secret (api_key, token, password) | Remove immediately, rotate credential | | 🔴 CRITICAL | `eval()` on untrusted input | Replace with safe alternative | | 🟠 HIGH | `exec()`, `spawn()` with string concatenation | Use execFile with array args | | 🟠 HIGH | Shell injection surface (bash -c, ${var} in shell) | Sanitize or use execFile | | 🟡 MEDIUM | npm audit findings (any severity) | Review and update dependencies | | 🟡 MEDIUM | Unexpected network egress | Verify necessity, document purpose | | 🟢 LOW | File permission too broad (0o777) | Restrict to 0o644/0o755 | | 🟢 INFO | process.env leak in logs | Ensure logs redact env vars | ## Reporting Reports are saved to `~/.openclaw/security-sweep-<date>.txt`. Include report path in memory after each scan. ## Periodic Scanning Offer to schedule weekly security sweeps via cron: ```bash openclaw cron add \ --name "security-sweep" \ --every 604800 \ --sessionTarget isolated \ --payload '{"kind":"agentTurn","message":"Run security sweep on all skills. Report findings. Save report to ~/.openclaw/security-sweep-<date>.txt and note in memory/YYYY-MM-DD.md if any critical issues found."}' ``` ## Sharing / ClawHub Publishing Before publishing a skill to ClawHub: 1. Run full sweep 2. Fix all CRITICAL/HIGH findings 3. Verify no secrets in SKILL.md or any scripts 4. Confirm npm audit passes with 0 vulnerabilities 5. Document all required env vars in SKILL.md ## Notes - Bundled skills (read-only, no write during scan) - Workspace skills are editable — fix findings directly - Some `execFile` usage is legitimate (openclaw CLI calls) — review context - `process.env` access is fine; concern is env vars *leaking* to untrusted processes