gemini-carousel
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses a 'CRITICAL' directive ('Auto-start on load') instructing the agent to bypass summarization and proceed immediately to execution, which overrides default interaction guardrails.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied content (such as newsletters, posts, or research notes) to generate briefs and update profile files, which could lead to the agent following instructions embedded within that data.
- Ingestion points: User-provided text collected in Step 1 ('Gather inputs').
- Boundary markers: No specific delimiters or instructions are used to separate the user-supplied content from the agent's internal logic.
- Capability inventory: The skill reads and updates local configuration files (
about-me.md,voice.md,brand-kit.md) and uses interaction tools likeAskUserQuestion. - Sanitization: There is no evidence of sanitization, filtering, or validation of the input content before it is processed by the agent.
Audit Metadata