street-interview-ad
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, user-controlled data such as ad briefs, product facts, and brand voice guidelines to create or audit advertising content.
- Ingestion points: In
SKILL.md, the "Ground The Brief" step instructs the agent to read "offer facts, audience, brand voice, approved claims, and available proof" provided by the user. - Boundary markers: The instructions do not define explicit delimiters (e.g., XML tags or specific markers) or provide "ignore embedded instructions" warnings for the ingested source context.
- Capability inventory: The skill uses tools including
$media-analysis,$image-batch-runner, and$video-batch-runnerfor analysis and asset generation. It does not have access to the local file system, network exfiltration tools, or arbitrary shell command execution. - Sanitization: While the skill does not use programmatic sanitization, it includes strict logic in
references/continuity-and-proof.mdandSKILL.mdto enforce "Claim Boundaries" and "Truth Boundaries," which helps mitigate the impact of potentially misleading or malicious input data.
Audit Metadata