dispatching-agent-work
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill implements robust activation checks, explicitly instructing the agent not to trigger orchestration modes when the skill is merely being reviewed or edited. It differentiates between one-time requests and persistent 'dispatch mode' to prevent unintended behavioral changes.
- [DATA_EXPOSURE]: Includes explicit safety warnings for the agent to avoid placing credentials, secrets, or private data into URI-encoded deep links used for task creation.
- [INDIRECT_PROMPT_INJECTION]: As a task orchestrator, the skill identifies the attack surface of ingesting untrusted user data. It mitigates this by instructing the agent to define clear 'dispatch contracts' with boundaries on authority and prohibited actions, and by requiring the removal of any backends or capabilities for which authorization is not present.
- [COMMAND_EXECUTION]: The skill directs the agent to utilize host-native operations for task management rather than attempting to emulate orchestration through shell commands or undocumented system calls.
Audit Metadata