write-korean-technical

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill uses the uv run command to execute Python scripts (lint.py, outline.py, terms.py) from a sibling directory named natural-korean. The execution path is dynamically resolved by the agent based on the provided instructions.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external code and log files to generate technical documentation, which introduces an indirect prompt injection surface.
  • Ingestion points: The agent is instructed to read actual files, logs, and command outputs from the environment (SKILL.md).
  • Boundary markers: No specific boundary markers are defined for the ingested data, although the quality gate (references/quality-gate.md) requires verification of code and commands.
  • Capability inventory: The skill uses shell commands via uv run to perform linting and analysis on the generated documents (SKILL.md).
  • Sanitization: There are no explicit instructions to sanitize or escape the content of the ingested logs or code before processing.
  • [COMMAND_EXECUTION]: The skill invokes local command-line tools for linting, outline verification, and terminology checks during the document creation process.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 11:17 PM
Security Audit — agent-trust-hub — write-korean-technical