QA测试主管 测试策略治理
AdvisoryAudited by Static analysis on May 8, 2026.
Overview
No suspicious patterns detected.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The agent may follow additional local project instructions that are not visible in this review.
The skill depends on project-local reference files that were not included in the submitted artifact set. This is normal for a project-specific QA skill, but those files may influence actual behavior.
Source Material - `AI-ENTRY.md` - `CLAUDE.md` - `dev/ai/skills/testing/SKILL.md`
Before using the skill, confirm the referenced project guidance files are trusted and consistent with your intended QA workflow.
When WLS validation is needed, the workflow may involve starting and stopping a local test instance.
The skill may lead the agent to operate a dedicated test runtime for validation. The instruction is scoped to WLS validation and includes cleanup, so it appears purpose-aligned.
Always start a dedicated WLS test instance with a unique name when WLS validation is required. - Always stop the AI test instance after testing.
Confirm WLS validation is necessary, use an isolated non-production instance, and verify it is stopped after testing.
QA blockers or risk details may be routed to a named technical lead role during use.
The skill includes an internal escalation instruction to a named Weline role or agent. No external endpoint or hidden channel is shown, but issue details could be shared as part of collaboration.
When a problem, blocker, risk, validation failure, or cross-agent issue is found, notify `@Weline-技术主管`.
Ensure any escalation stays within the intended workspace and does not include secrets or sensitive user data unless explicitly approved.