check-consistency

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Python scripts (scripts/validate_glossary.py, scripts/term_read.py, scripts/term_edit.py) via the uv package manager. This is standard behavior for a project-specific automation tool.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes documentation content located in docs/src/content/docs/. This represents a surface for indirect prompt injection if the documentation being scanned contains malicious instructions intended to mislead the AI agent during the terminology audit process.
  • Ingestion points: Documentation files and glossary files.
  • Boundary markers: None explicitly defined in the prompt instructions to isolate data from instructions.
  • Capability inventory: File system read/write via local scripts.
  • Sanitization: Not explicitly mentioned; the skill relies on the logic within the Python scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 06:03 PM
Security Audit — agent-trust-hub — check-consistency