external-cannbot-ops-pypto-op-perf-tune

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires shell commands for NPU diagnostics, environment variable configuration, and execution of local Python testing scripts. These are standard operations for hardware-accelerated operator tuning.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted performance data.
  • Ingestion points: The analyze_perf.py script reads from bubble_analysis.log.
  • Boundary markers: None identified.
  • Capability inventory: Includes command execution and local file writing.
  • Sanitization: Employs strict regex patterns to extract numeric values, mitigating input risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 03:54 PM
Security Audit — agent-trust-hub — external-cannbot-ops-pypto-op-perf-tune