Skip to content

Security: ch4p-labs/ch4p

Security

SECURITY.md

Security Policy

Supported versions

Version Supported
0.1.x Yes

Reporting a vulnerability

Do not open a public GitHub issue for security vulnerabilities.

Please use GitHub's private vulnerability reporting to disclose security issues responsibly. This keeps the report confidential until a fix is available.

What to include

  • Description of the vulnerability
  • Steps to reproduce
  • Affected versions and components
  • Impact assessment (if known)

Response timeline

  • Acknowledgement — within 48 hours
  • Initial assessment — within 1 week
  • Fix or mitigation — depends on severity, but we aim for patches within 30 days for critical issues

Security architecture

ch4p ships with nine defense layers enabled by default:

  1. Filesystem scoping with symlink escape detection
  2. Command allowlist with shell metacharacter blocking
  3. AES-256-GCM encrypted secrets with PBKDF2 key derivation
  4. Output sanitization (25 regex patterns)
  5. Input validation (prompt injection and data exfiltration detection)
  6. Autonomy levels (readonly / supervised / full)
  7. SSRF protection (private IP blocking, DNS checks, cloud metadata guards)
  8. Secure file permissions (transcripts 0o600, log dirs 0o700)
  9. Pairing token expiration (30-day TTL)

For full details, see:

There aren’t any published security advisories