wait
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its messaging interface.
- Ingestion points: The
hive msg subandhive msg inboxcommands (SKILL.md) read message content from topics that could be controlled by other agents or external actors. - Boundary markers: The documentation does not implement delimiters or instructions to ignore embedded commands in received messages.
- Capability inventory: The skill examples demonstrate using these messages for flow control and shell-based logic via the
hivetool. - Sanitization: No methods for sanitizing, escaping, or validating the content of received messages are described.
Audit Metadata