handover
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data to generate continuity instructions for future agent sessions, creating a potential injection vector.
- Ingestion points: The skill ingests 'Current objective', 'Relevant artifact paths' (PRDs, slices, commits), and 'Ordered runbook commands' from the current workspace context (SKILL.md).
- Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded prompts within the ingested artifacts when generating the handover document.
- Capability inventory: The skill utilizes shell command execution (
wikiCLI,git,mktemp) and file system writes to save the handover record. - Sanitization: There is no evidence of sanitization or filtering of external content before it is interpolated into the handover summary for the next agent.
Audit Metadata