Security audit
Security checks for vulnerabilities and agentic risk
The skill is a coherent task tracker, but it is too eager to execute broadly detected or ambiguous task lists and persist them locally.
Install only if you want the agent to aggressively turn compound messages into executable task queues. Avoid using it on prompts that contain secrets, personal data, or quoted instructions you only want analyzed; review and delete TASK_QUEUE.md when it is no longer needed.
The activation logic is intentionally over-broad: it triggers on 'ANY' message with 3+ actionable items and even says to activate when uncertain. In a security-sensitive agent, this can cause the skill to intercept benign or ambiguous user input, restructure it, and begin persistence/execution workflows without clear user consent, increasing the chance of unintended actions and data handling.
The skill instructs the agent to write user-derived task content to workspace/TASK_QUEUE.md before execution and to keep using that file as the source of truth, but it never requires notifying the user or minimizing what is stored. This creates a privacy and data-retention risk because sensitive instructions, credentials, personal data, or internal project details from compound prompts may be persisted to disk unnecessarily.
No suspicious patterns detected.
