multitask-coordinator
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and associated reference files follow security best practices for AI agent coordination. No malicious patterns were detected.
- [PROMPT_INJECTION]: The skill is focused on task scheduling and coordination. It does not contain jailbreak attempts, safety filter bypasses, or instructions to ignore previous constraints.
- [DATA_EXFILTRATION]: There is no evidence of data exfiltration. The ephemeral-shared-memory.md reference explicitly prohibits the storage of credentials, secrets, or capability URLs in coordination files.
- [REMOTE_CODE_EXECUTION]: The skill does not perform any remote code downloads or execution. All task execution is managed through defined local agent workers.
- [COMMAND_EXECUTION]: The skill uses environment inspection tools like git status for context, but it does not facilitate arbitrary or unsafe command execution. Cleanup operations are restricted to verified, owned temporary directories.
Audit Metadata