llm-usage-researcher

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted 'test inputs' (like external code snippets or documentation) to evaluate the performance of different LLM models.\n
  • Ingestion points: Test inputs are gathered from the user and processed in the evaluation loop (as defined in SKILL.md).\n
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings for the data being evaluated.\n
  • Capability inventory: The skill performs network operations via LLM API calls and writes multiple files to the local file system (dashboard.html, comparison.tsv, results.json).\n
  • Sanitization: There is no evidence of sanitization or escaping of the external content before it is interpolated into the prompts for the target models.\n- [SAFE]: The generated dashboard.html includes references to Chart.js from a public CDN. This is a well-known and safe practice for providing interactive data visualization in generated reports.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 03:29 AM