handoff

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by summarizing conversation history and user arguments for use by a subsequent agent.
  • Ingestion points: Conversation history and user-provided arguments (via SKILL.md).
  • Boundary markers: Absent. The instructions do not require the agent to use delimiters or warnings for the summarized content.
  • Capability inventory: File system write (saving to the OS temporary directory).
  • Sanitization: The skill contains an instruction to redact sensitive data (API keys, passwords, PII), but lacks automated sanitization or validation of the summarized content to prevent instruction carry-over.
  • [DATA_EXFILTRATION]: The skill is explicitly instructed to save handoff documents to the user's OS temporary directory instead of the current workspace. While this is the intended functionality, it involves writing potentially sensitive summary data to locations outside the controlled project environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 12:23 AM