influencer-campaign-brief
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of instructional Markdown and evaluation data with no executable logic, shell commands, or network requests.
- [INDIRECT_PROMPT_INJECTION]: The skill processes campaign and creator data from user input, which presents a surface for indirect instructions within a produced brief.
- Ingestion points: User input is collected through the campaign interview process defined in SKILL.md.
- Boundary markers: The skill uses a Quality Gate (Step 10) and validation steps to ensure the output aligns with the user's requirements and signed deal terms.
- Capability inventory: The skill's operations are restricted to text generation; it has no access to shell, file system, network, or external APIs.
- Sanitization: Manual reconciliation against the 'signed deal' serves as a verification layer to prevent the adoption of untrusted data into deliverable specifications.
- [SAFE]: References to other skills (e.g., mbfinotti/partnerships-skills) are confirmed as vendor-specific resources belonging to the author and do not present third-party supply chain risks.
Audit Metadata