python-performance-optimization

Pass

Audited by Gen Agent Trust Hub on May 10, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to process and analyze user-provided Python scripts to identify performance bottlenecks. This constitutes an indirect prompt injection surface.
  • Ingestion points: User-supplied code blocks referenced in the triggers in SKILL.md.
  • Boundary markers: Not explicitly defined in the skill content.
  • Capability inventory: Shell command execution for profiling tools (cProfile, py-spy) and network access via requests and aiohttp for I/O benchmarking.
  • Sanitization: No explicit code validation or filtering is described, relying on the agent's environment security.
  • [COMMAND_EXECUTION]: Includes standard shell commands for installing and running industry-standard profiling utilities via pip and uv tools.
  • [EXTERNAL_DOWNLOADS]: Includes code examples that perform network requests to httpbin.org, a well-known service for testing HTTP operations.
Audit Metadata
Risk Level
SAFE
Analyzed
May 10, 2026, 03:15 AM
Security Audit — agent-trust-hub — python-performance-optimization