case-study-builder

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data provided by the user (client engagement facts) and interpolates it directly into a prompt template.
  • Ingestion points: Raw user inputs for [NAME/TYPE], [WHAT THEY STRUGGLED WITH], [WHAT WAS IMPLEMENTED], and [MEASURABLE OUTCOMES] (SKILL.md).
  • Boundary markers: None identified. The skill lacks delimiters (e.g., XML tags or triple quotes) or explicit instructions to the agent to treat the input solely as data.
  • Capability inventory: The skill is restricted to text generation and does not have access to file-writing, network, or subprocess tools.
  • Sanitization: None identified. There is no logic to filter or escape instructions embedded within the user data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 08:55 AM
Security Audit — agent-trust-hub — case-study-builder