api-patterns

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFE
Full Analysis
  • COMMAND_EXECUTION (SAFE): The script scripts/api_validator.py performs static analysis of local project files using Python's standard library (pathlib, re, json). It does not invoke system shells or execute the code it analyzes.\n- DATA_EXFILTRATION (SAFE): There are no network modules or commands (e.g., requests, curl, socket) present in the skill's code or documentation.\n- INDIRECT_PROMPT_INJECTION (LOW): The validator script is an ingestion surface as it reads and analyzes user-provided project files.\n
  • Ingestion points: scripts/api_validator.py reads files via Path.read_text at lines 43 and 85.\n
  • Boundary markers: None; the script's output does not use delimiters to separate ingested file data (such as API paths or JSON keys) from the agent's instructions.\n
  • Capability inventory: The script is limited to local file system read operations for specific file extensions. It has no capabilities for file writing, network access, or process execution.\n
  • Sanitization: None; the script may include keys or values from analyzed JSON files directly in its printed output, but the risk is negligible given the restricted scope of the tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 04:11 AM