handover
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a handover document by aggregating data from various untrusted sources, which could potentially contain malicious instructions intended to influence the next agent session.
- Ingestion points: The skill reads the entire conversation history, plan files, memory files, and content from GitHub issues and pull requests (SKILL.md).
- Boundary markers: The output template provides structure but does not include explicit boundary markers or instructions to the next agent to ignore embedded commands within the summarized content.
- Capability inventory: The skill utilizes file system access to write to temporary directories and uses GitHub API integration to post comments and manage labels.
- Sanitization: The skill contains an explicit instruction to redact credentials and PII (Personal Identifiable Information), which is a positive security control. However, it does not include measures to sanitize or escape executable natural language instructions that might be present in the source data.
Audit Metadata