Summary

Following the CLAWDINATOR guidance to open an issue first before submitting security-related PRs. This proposal bundles 4 complementary security hardening improvements that reduce supply-chain and runtime risks.

Motivation

OpenClaw has 500k+ lines of code across multiple platforms with browser automation, command execution, and WebSocket control plane. While the project already has excellent security foundations (detect-secrets, openclaw security audit, loopback-by-default), there are baseline gaps that could be closed with minimal risk:

1. Remote asset dependency (Control UI loads SVG from external CDN)
2. Missing HTTP security headers (no baseline CSP/XFO/nosniff on gateway responses)
3. Unused security tooling configs (actionlint + zizmor configs exist but aren't enforced in CI)
4. No dependency vulnerability scanning (no CodeQL or dependency-review in CI)

Proposed Changes

Phase 1: Supply Chain Hardening (docs + CI)

1. Localize Control UI logo asset (addresses #5170)

• Replace hardcoded mintcdn.com SVG URL with bundled local asset
• PR reference: #6166 (closed), #6161 (alternative implementation)
• Risk reduced: Removes external CDN dependency that could be compromised

2. Wire up existing CI security configs

• Add .github/workflows/actionlint.yml (uses existing .github/actionlint.yaml)
• Add .github/workflows/zizmor.yml (uses existing zizmor.yml)
• PR reference: #6194 (closed, CI was fixed)
• Risk reduced: Catches workflow security footguns and misconfigurations

Phase 2: Runtime + Dependency Hardening

3. Add baseline HTTP security headers

• Add conservative headers to gateway HTTP responses: X-Content-Type-Options: nosniff, Referrer-Policy, X-Frame-Options: SAMEORIGIN, Permissions-Policy
• Intentionally scoped to be safe for both Control UI and API routes
• PR reference: #6171 (closed, had review feedback incorporated)
• Risk reduced: Reduces clickjacking, content-sniffing, and feature abuse surface

4. Add dependency scanning

• Add .github/workflows/dependency-review.yml (GitHub official action)
• Add .github/workflows/codeql.yml (JavaScript/TypeScript analysis)
• PR reference: #6195 (closed, CI was green)
• Risk reduced: Catches vulnerable dependencies and common insecure code patterns

Implementation Notes

• All changes are additive - no breaking changes to existing functionality
• Minimal permissions - CI workflows use contents: read only (CodeQL needs security-events: write)
• Follows existing patterns - Uses same action versions and config styles as current CI
• Tested - All PRs were validated with pnpm build, pnpm lint, and targeted tests

Request for Maintainers

@steipete (or active maintainers) - would you be open to these 4 hardening PRs? If yes, I can:

1. Re-open each PR individually with reference to this issue
2. Or combine into a single "security hardening bundle" PR if preferred
3. Or adjust scope based on your feedback

The goal is to reduce "death by a thousand cuts" security risks while keeping the PRs small and reviewable.

References

• Security docs: https://docs.openclaw.ai/gateway/security
• Related issue: [Bug]: Possible security problem with the .svg file #5170 (remote SVG)
• Closed PRs with full implementation: security(ui): remove remote mintcdn asset reference #6166, security(gateway): add baseline HTTP security headers #6171, ci(security): add actionlint + zizmor workflows #6194, ci(security): add dependency review + CodeQL #6195