Skip to content

Security: OpenSecBench/ActionSieve

.github/SECURITY.md

Security Policy

Reporting a Vulnerability

If you discover a security vulnerability in ActionSieve, please report it via GitHub's private vulnerability reporting. Do not open a public issue.

You should receive an acknowledgment within 48 hours.

Scope

ActionSieve is a static analysis tool that parses CI/CD pipeline definitions. Security-relevant areas include:

  • YAML parsing — must use safe_load only, never yaml.load
  • File handling — size limits, path traversal prevention
  • No code execution — the scanner must never execute code from scanned repos
  • No repo-controlled config — scanned repos cannot suppress findings without explicit --trust-repo-profile

Supported Versions

Version Supported
latest Yes

Security Design Principles

  • No eval(), exec(), or subprocess with shell=True on user input
  • No importing or executing code from scanned repositories
  • File size caps on parsed files (1MB per workflow)
  • Recursion depth limits for composite action resolution
  • Reference CI workflows use pull_request (not pull_request_target), pin all action refs to SHA

There aren't any published security advisories