worker
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill documents a procedural framework for subagent lifecycle management, including identity parsing, startup ACKs, and evidence-backed completion. No malicious code, obfuscation, or unauthorized access to sensitive system resources was detected.
- [PROMPT_INJECTION]: The protocol establishes an indirect prompt injection surface by directing the agent to read and follow instructions from shared team artifacts.
- Ingestion points: The agent is instructed to read instructions from
.omw/team/<teamName>/mailbox/<workerName>.mdand.omw/team/<teamName>/tasks/task-<id>.json. - Boundary markers: The instructions do not specify boundary markers or provide guidance on treating the content of these files as potentially untrusted data.
- Capability inventory: As a worker subagent, the model is expected to perform general tasks which include file system operations and command execution within the workspace.
- Sanitization: The protocol lacks explicit sanitization or validation requirements for instructions received via shared files.
Audit Metadata