influencer-measurement-framework

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data during its interview phase to construct measurement plans, creating a potential surface for indirect instructions.
  • Ingestion points: User responses collected in the Interview section of SKILL.md.
  • Boundary markers: None identified.
  • Capability inventory: No executable code, shell access, network operations, or sensitive file system writes are present in SKILL.md or any referenced markdown files.
  • Sanitization: No explicit sanitization or filtering of user-provided strings is documented.
  • [SAFE]: The skill is entirely instructional and document-based, referencing established industry guidelines (ANA, AMEC, IAB) and open-source analytics tools without including malicious patterns, obfuscation, or credentials.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 09:17 PM
Security Audit — agent-trust-hub — influencer-measurement-framework