nodumb-loop
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a logic layer for routing technical tasks, which involves the ingestion and processing of untrusted user-supplied data.\n
- Ingestion points: User-provided product or technical task descriptions are processed in SKILL.md to determine the appropriate routing layer.\n
- Boundary markers: Absent; there are no instructions or delimiters provided to the agent to distinguish between task data and system instructions, or to ignore embedded instructions in the input.\n
- Capability inventory: The skill routes the agent to execute other specialized skills including 'nodumb', 'edge-hunt', and 'ask-nodumb', which typically involve code analysis, fact verification, and external interaction.\n
- Sanitization: Absent; the skill does not specify any validation, escaping, or filtering for the content of the tasks it routes.
Audit Metadata