handoff

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): No attempts to override system prompts or bypass safety filters were detected. The instructions are focused on structured documentation tasks.
  • [Data Exposure & Exfiltration] (SAFE): The skill only interacts with a local 'HANDOFF.md' file within the project directory. There are no network calls or attempts to access sensitive system paths (e.g., SSH keys or environment variables).
  • [Indirect Prompt Injection] (LOW): A surface for indirect prompt injection exists because the skill reads content from an existing 'HANDOFF.md' file which could contain malicious instructions.
  • Ingestion points: Reads 'HANDOFF.md' in Step 2.
  • Boundary markers: Absent; the content of the file is read without specific delimiters or instructions to ignore embedded commands.
  • Capability inventory: Limited to reading and writing the 'HANDOFF.md' file; the skill does not have access to the shell, network, or other sensitive tools.
  • Sanitization: None; the skill reads and processes the file content directly.
  • [Persistence Mechanisms] (SAFE): While the skill creates a persistent file, it is for documentation purposes and does not attempt to modify startup scripts, crontabs, or system services.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:30 PM