write-readme

Pass

Audited by Gen Agent Trust Hub on Oct 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from project files to generate a README, creating a potential attack surface for indirect instructions.
  • Ingestion points: The agent is instructed to read the existing README.md, project entry points, examples, tests, and build metadata files.
  • Boundary markers: There are no instructions providing boundary markers or explicit guidelines to ignore instructions embedded within the files being read.
  • Capability inventory: The skill generates Markdown documentation and may execute code examples found in the project.
  • Sanitization: No sanitization, escaping, or filtering of the ingested content is specified.
  • [DYNAMIC_EXECUTION]: The instructions direct the agent to execute code snippets found in the analyzed project to verify their correctness.
  • Evidence: The skill explicitly tells the agent to "run examples when practical and safe" and to report which examples could not be executed. This involves running arbitrary code from the target repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 29, 2026, 08:54 AM