firecrawl-shop
Pass
Audited by Gen Agent Trust Hub on Jul 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of ingesting and processing data from external, untrusted sources on the open web.
- Ingestion points: The skill uses Firecrawl to search and scrape reviews, product pages, specifications, pricing, and forums (e.g., Reddit).
- Boundary markers: The instructions do not define boundary markers or explicit commands for the agent to ignore instructions embedded within the retrieved web content.
- Capability inventory: The skill allows the agent to navigate shopping sites and perform cart actions (adding items) when an authenticated session is available.
- Sanitization: No sanitization, filtering, or validation of the content retrieved from external sources is described in the workflow.
Audit Metadata