diataxis-writer

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes a utility script scripts/classify-doc.sh which performs keyword frequency analysis on text files to help determine the Diataxis category. The skill instructions guide the agent to execute this script locally using bash. This is a controlled execution of a static, distributed script for the skill's primary purpose.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided documentation, which constitutes an untrusted data surface.
  • Ingestion points: The agent reads document content provided by the user in the prompt or via file paths passed to the classification script.
  • Boundary markers: The instructions do not explicitly specify XML tags or delimiters for the ingested documentation content, relying on the agent's context processing.
  • Capability inventory: The skill has the capability to read files and execute a local bash script (scripts/classify-doc.sh) to perform analysis.
  • Sanitization: The analysis script performs keyword matching and does not execute the content of the processed documents, limiting the risk of exploitation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 11:04 AM
Security Audit — agent-trust-hub — diataxis-writer