agent-readiness

Pass

Audited by Gen Agent Trust Hub on Oct 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill's core script, scripts/probe.py, uses subprocess.run to execute git commands. This functionality is restricted to the git binary and is used to retrieve tracked file lists and commit history for auditing purposes. Additionally, the test suite in tests/test_cli.py executes the probe script via subprocess.run to verify CLI behavior.
  • [INDIRECT_PROMPT_INJECTION]: The skill audits third-party repositories, creating a surface for indirect prompt injection where malicious instructions in the audited code could influence the agent's evaluation.
  • Ingestion points: scripts/probe.py reads file contents (via _read_bounded) and git history (via git log).
  • Boundary markers: None implemented in the probe output; the skill relies on instructions in SKILL.md to scope agent judgements.
  • Capability inventory: The skill can execute local git commands and read local files. It has no network access or file-writing capabilities.
  • Sanitization: Minimal; the script decodes file bytes with errors="replace" to handle invalid UTF-8 sequences.
  • [DYNAMIC_EXECUTION]: The test configuration in tests/conftest.py utilizes importlib.util.spec_from_file_location and exec_module to load scripts/probe.py as a module for automated testing. This is a standard and safe practice for Python test runners.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 16, 2026, 05:06 PM