handoff
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill summarizes conversation history and references external artifacts to create a handoff for subsequent agent sessions, creating a surface where malicious instructions could influence future agents.\n
- Ingestion points: The skill ingests the "current conversation" and references external artifacts like specs, plans, ADRs, and issues by path or URL (SKILL.md).\n
- Boundary markers: There are no instructions to use delimiters or to explicitly ignore instructions embedded within the conversation history or referenced artifacts.\n
- Capability inventory: The skill instructs the agent to perform file-write operations to the operating system's temporary directory and generates a "suggested skills" section that influences the next agent's tool usage (SKILL.md).\n
- Sanitization: The skill includes explicit instructions to redact sensitive information such as API keys and PII, which provides protection against data exposure but does not specifically sanitize against prompt injection.
Audit Metadata