trend-spotter
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acknowledges the risk of processing untrusted data and includes strong defensive instructions to mitigate it.
- Ingestion points: The skill processes scraped page text, reviews, comments, emails, and third-party exports as noted in SKILL.md.
- Boundary markers: Present. The skill contract and data source sections explicitly instruct the agent to treat these inputs as evidence only and to ignore any embedded instructions.
- Capability inventory: No executable scripts are included in the skill. Tool use is limited to those provided by the platform (e.g., search, marketing connectors).
- Sanitization: Relies on instructional guardrails to ensure external content does not override the agent's primary mission.
- [DATA_EXFILTRATION]: The references provide best practices for managing external service connections, specifically instructing that API keys must be kept out of skill files and source control. It also defaults to read-only behavior for scripts to prevent unauthorized modifications.
Audit Metadata