oneos-qa-verify
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a structured manual QA process. It includes explicit safety prohibitions against automated deployments or state changes without human oversight, such as the rule that 'Release Candidate' status is a suggestion for a human, not an authorization for the AI to publish.- [DATA_EXPOSURE]: The skill references other skill files located at
~/.cursor/skills/YunxiaoQA/SKILL.mdand relative directories (e.g.,../oneos-biz-rules/SKILL.md). These references are used to maintain consistency across the 'oneos' skill suite and do not target sensitive system configuration files or credentials.- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data such as test evidence and external product requirements. - Ingestion points: Data enters the context via the 'Evidence Template' and referenced requirement scripts from other skill files.
- Boundary markers: The skill uses markdown headers and a mandatory structure (e.g., '## 测试证据') to separate instructional context from user-provided data.
- Capability inventory: The skill primarily performs context aggregation from local skill files; it does not list or utilize high-risk tools for command execution, network exfiltration, or file modification.
- Sanitization: No explicit sanitization or filtering of input content is specified in the instructions.
Audit Metadata