revyl-mcp-analyze
Analyze failed Revyl MCP test executions and classify them as real bugs, flaky tests, infra issues, or test-design improvements.
来源信息
- 仓库
- RevylAI/revyl-cli
- 最近来源活动
- 2026年9月8日 03:59
- 检测到的 SKILL.md 语言
- 英语
- 星标
- 520
- 分支
- 26
安装方式
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
检查来源文件
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
文件资源管理器
2 个文件正在显示 SKILL.md
SKILL.md
来源说明 · 只读预览- name
- revyl-mcp-analyze
- description
- Analyze failed Revyl MCP test executions and classify them as real bugs, flaky tests, infra issues, or test-design improvements.
# Revyl MCP Analyze Skill
## Evidence Flow
1. Start from task ID after `run_test`.
2. Pull status payload:
- `get_test_status(task_id="...")`
3. Inspect failed step expectations and observed reasoning.
4. Classify:
- REAL BUG
- FLAKY TEST
- INFRA ISSUE
- TEST IMPROVEMENT
5. Recommend exact next tool/action.
## Classification Rules
- REAL BUG: actions complete but app behavior contradicts expected outcome.
- FLAKY TEST: app behavior is acceptable but validation is too brittle.
- INFRA ISSUE: no meaningful app steps executed or setup/device/build failure.
- TEST IMPROVEMENT: test structure is weak or allows ambiguity.
## Required Output
```text
Task: <task-id>
Classification: <REAL BUG | FLAKY TEST | INFRA ISSUE | TEST IMPROVEMENT>
Confidence: <HIGH | MEDIUM | LOW>
Evidence:
- Expected: <summary>
- Observed: <summary>
- Why: <short rationale>
Next action:
- <specific update_test / rerun / bug filing action>
```
在 GitHub 查看