instagram-trend-scout
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for social media research and uses the
web_searchtool to collect public information. It correctly handles input validation and requires no authentication or access to private data. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the public web via search results to identify trends. While this presents an indirect prompt injection surface, it is the primary purpose of the skill and the risk is mitigated by the skill's limited capabilities, as it only returns structured data and does not perform file operations or network exfiltration.
- Ingestion points: Web search results (Step 3) are used to collect content angles, hooks, and engagement signals.
- Boundary markers: The skill does not explicitly define delimiters for the external content but instructs the agent to normalize data into a specific JSON schema.
- Capability inventory: The skill utilizes the
web_searchtool; it lacks capabilities for writing to the filesystem, executing arbitrary code, or performing secondary network requests. - Sanitization: Content is normalized and summarized into structured fields, which limits the potential impact of embedded instructions in the search results.
Audit Metadata