do-redteam
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process and analyze external arguments, strategies, and plans, which introduces a potential attack surface for indirect prompt injection.
- Ingestion points: External data enters the agent's context through placeholders such as
[Problem/Task description]inWorkflows/AdversarialValidation.mdand[Full original argument]inWorkflows/ParallelAnalysis.md. - Boundary markers: The workflows use
<instructions>tags to define agent behavior, but user-supplied content is appended outside these tags without secure delimiters or escaping, which could allow an attacker to attempt to override the agent's primary mission. - Capability inventory: The skill is limited to text generation, critique, and synthesis using persona-based sub-agents. It does not utilize file-system writes, network operations, or arbitrary code execution.
- Sanitization: The skill lacks explicit sanitization, validation, or filtering logic to neutralize malicious instructions embedded within the arguments it is tasked to analyze.
Audit Metadata