context-engineering

Pass

Audited by Gen Agent Trust Hub on Apr 4, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill defines trust levels for context and explicitly labels user-submitted content and third-party API responses as 'Untrusted', instructing agents to treat instruction-like text in these sources as data rather than directives.
  • [COMMAND_EXECUTION]: Examples of standard development commands (e.g., npm build, npm test) are provided for context configuration; these are common project tasks and do not represent arbitrary or dangerous code execution.
  • [DATA_EXFILTRATION]: The skill advocates for establishing clear boundaries by instructing agents via rules files to never commit secrets or environment variables, which helps prevent accidental data exposure.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 4, 2026, 10:17 PM