handoff
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process the current conversation history to generate a handoff artifact. This creates a surface for indirect prompt injection if the conversation contains instructions intended to manipulate the behavior of the recipient agent.
- Ingestion points: The entire session history is used as input in
SKILL.md. - Boundary markers: The skill does not define specific delimiters or guardrails to prevent embedded instructions from being treated as authoritative by the next agent.
- Capability inventory: The skill itself performs text summarization and file writing; it does not request network access, subprocess execution, or unsafe dynamic execution.
- Sanitization: The instructions do not specify sanitization for prompt injection patterns, although they do mandate the removal of credentials.
- [DATA_EXPOSURE]: The skill includes a specific directive for content discipline, requiring that keys, passwords, and personal identity information (PII) must be excluded from the handoff documents. This is a positive security practice for session management.
Audit Metadata