payloads-cache-deception
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion and analysis of untrusted external data, which creates a potential vector for indirect prompt injection if the agent encounters malicious instructions embedded in target system responses.
- Ingestion points: The workflow in
SKILL.mdrequires the agent to identify injection points, enumerate endpoints, and capture PoC evidence including HTTP logs and curl outputs. - Boundary markers: There are no specific instructions or delimiters provided to help the agent distinguish between its own operational instructions and potentially malicious data within the captured logs or target responses.
- Capability inventory: The skill encourages the use of network utilities (
curl) and references integration with analysis tools (REA MCP) to perform active testing and verification inSKILL.md. - Sanitization: No sanitization, filtering, or validation mechanisms are defined for processing the external data gathered during reconnaissance and exploitation phases.
Audit Metadata