Agent Task Checklist Management
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill mandates the creation and frequent updating of a
task.mdfile to track project steps. This workflow creates an entry point for untrusted data if project descriptions or requirements are sourced externally and automatically converted into tasks. - Ingestion points: The
task.mdfile (referenced in SKILL.md) is the primary location where project-related instructions are stored and processed. - Boundary markers: The instructions lack specific delimiters or guidance to treat task content as data rather than executable instructions.
- Capability inventory: The agent performs file-write operations to maintain the task list.
- Sanitization: No sanitization or validation logic is defined for the strings used to populate the task list.
Audit Metadata