handoff
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: No malicious code, obfuscation, or unauthorized network activity was detected. The skill is designed to improve agent workflow efficiency.
- [COMMAND_EXECUTION]: The skill writes conversation summaries to a file named
handoff.md. The target path is determined based on user input, project configuration files, or a default directory within~/.agent-skills/changes/. While this involves file system interaction, it is a core functional requirement and follows explicit resolution logic. - [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it summarizes data from the current conversation which could contain untrusted input. However, the skill includes guardrails instructing the agent to focus only on current context and specified aspects, which mitigates the risk of capturing or acting upon malicious instructions embedded in the dialogue.
Audit Metadata