Skills
Navigate
skills/xsyetopz/skills/remove-unneeded-compatibility-code/Gen Agent Trust Hub

remove-unneeded-compatibility-code

Pass

Audited by Gen Agent Trust Hub on Sep 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted project source code, which could contain malicious instructions designed to manipulate the agent's behavior during the cleanup process.
  • Ingestion points: Project source code files (Python, JavaScript, etc.) are parsed by the scripts/find_compat_python.py script and reviewed by the agent as part of the compatibility cleanup workflow.
  • Boundary markers: None identified.
  • Capability inventory: The skill has the capability to modify or delete files (during code cleanup) and execute local commands (scanner script and unit tests).
  • Sanitization: The scanner script uses static analysis via Python's ast module, which parses code into an Abstract Syntax Tree without executing it, mitigating direct execution risks during the scanning phase.
  • [COMMAND_EXECUTION]: The skill's workflow involves executing local scripts and testing frameworks to identify and verify code changes.
  • Evidence: The SKILL.md workflow and assets/examples/verify.sh script involve executing the bundled scripts/find_compat_python.py tool and running project unit tests using python3 -m unittest.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 27, 2026, 02:47 AM