proj-task
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill maintains project documentation by ingesting user requirements and task updates, creating a persistence surface for untrusted data.
- Ingestion points: User-provided requirements and status updates are used to populate and update task documents (found in SKILL.md under '初始化任务文档' and '更新任务状态').
- Boundary markers: The skill mandates a specific markdown structure defined in templates/task-doc.md, which helps delineate different data sections.
- Capability inventory: The skill has the capability to write and update markdown files within the project's
docs/directory hierarchy (SKILL.md). - Sanitization: There is no explicit instruction for the agent to sanitize or escape user-provided text before writing it to markdown files, which could lead to the storage of malicious instructions within the project documentation that may be interpreted during subsequent document reads.
Audit Metadata