aiseotracker
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external citations and brand snippets via its API tools, creating a vulnerability surface.
- Ingestion points: Data retrieved through tools like
aiseo_list_snippetsandaiseo_list_sources, as well as instruction fetches from the vendor domain. - Boundary markers: The instructions explicitly require the agent to 'Answer from tool JSON only' and not to invent data.
- Capability inventory: The agent can call various data-retrieval tools but is specifically barred from sending emails, posting content publicly, or initiating payments.
- Sanitization: There are no mentioned filters for external tool data.
- [EXTERNAL_DOWNLOADS]: The skill instructions direct the agent to fetch additional instructions and configurations from the vendor's website at
aiseotracker.com.
Audit Metadata