Skip to content

Security: Parad0x-Labs/openclaw-skills

SECURITY.md

Security Policy

Supported Security Posture

This repository is under active development. Security fixes are applied to the current main branch first.

Current notable protections:

  • LSEC v2 AEAD encryption (AES-256-GCM) with authenticated headers (AAD)
  • Fail-closed secret handling (no hardcoded fallback encryption secrets)
  • Path safety policy for OpenClaw/TraceVault wrappers (denylist + explicit risky override phrase)
  • Output permission hardening (0600 files / 0700 directories where supported)
  • CI benchmark regression gates and CLI JSON contract tests

Reporting a Vulnerability

Please do not open public issues for suspected security vulnerabilities.

Send a report to:

  • hello@parad0xlabs.com

Include:

  • affected file(s) / command(s)
  • proof of concept or reproduction steps
  • impact assessment (data disclosure, path traversal, integrity bypass, DoS, etc.)
  • environment details (OS, Python version)

We will acknowledge receipt and triage as quickly as possible.

Scope

High-priority reports include:

  • encryption/authentication bypasses
  • path traversal / symlink escape in pack/restore flows
  • secret leakage in logs, JSON output, or reports
  • denial-of-service inputs causing unbounded memory/CPU in default paths

Non-Goals / Known Constraints

  • Performance regressions are handled through benchmark gates, not this process
  • Legacy blob backward compatibility is intentionally not guaranteed (LSEC v2 only)
  • Third-party ecosystem plugins (not maintained in this repo) are out of scope unless the issue is in Liquefy CLI/API contracts

There aren't any published security advisories