backlog-manager
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from a local file, creating a surface for potential indirect prompt injection.
- Ingestion points: The skill reads content from docs/需求池.md during the categorization, organization, and screening phases.
- Boundary markers: Structured Markdown templates are employed to organize data, though there are no specific delimiters to signal the agent to ignore embedded instructions.
- Capability inventory: The skill's capabilities are limited to reading and writing the docs/需求池.md file; no other scripts or subprocesses are present.
- Sanitization: The workflow mandates that the AI confirms all interpretations and proposed file changes with the user before performing write operations.
Audit Metadata