predexon
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses instructional steering to command the agent to prioritize this skill over standard browser or web_fetch tools for specific domains, which attempts to override the agent's default decision-making logic.
- [INDIRECT_PROMPT_INJECTION]: The skill fetches and processes data from external prediction market venues which presents an attack surface for instructions embedded in data.
- Ingestion points: External data is ingested via multiple GET endpoints targeting
http://localhost:8402/v1/pm/as defined inSKILL.md. - Boundary markers: The instructions lack delimiters or warnings to the agent to treat the retrieved content as untrusted.
- Capability inventory: The skill provides extensive read access to market data, smart money analytics, and wallet identities, but does not define write or execution capabilities.
- Sanitization: No validation or sanitization requirements are specified for the external data before it is processed by the agent.
Audit Metadata