allure-testops-mcp
MCP server for Allure TestOps — lets an LLM agent (Claude Code, Cursor, Windsurf, etc.) inspect test quality without leaving the chat.
README
allure-testops-mcp
<!-- mcp-name: io.github.mshegolev/allure-testops-mcp -->
MCP server for Allure TestOps. Lets an LLM agent (Claude Code, Cursor, OpenCode, etc.) query projects, launches, test cases and test results through the Allure REST API.
Python, FastMCP, stdio transport.
Works with any Allure TestOps instance — SaaS qameta.io or self-hosted / on-prem. Designed with corporate networks in mind: configurable proxy bypass, optional SSL-verify toggle, API-token auth.
Design highlights
- Tool annotations — every tool is marked
readOnlyHint: True/openWorldHint: True. All 6 tools are read-only; MCP clients won't ask for confirmation. - Structured output on every tool — each tool declares a
TypedDictreturn type, so FastMCP auto-generates anoutputSchemaand every result carries bothstructuredContent(typed payload) and a pre-rendered markdown text block. - Structured errors — auth, 404, 403, 429, 5xx, missing-env errors converted to actionable messages (e.g. "Authentication failed — verify ALLURE_TOKEN has API scope").
- Pydantic input validation — every argument has typed constraints (ranges, lengths, literals) auto-exposed as JSON Schema.
- Pagination — list tools return a
paginationblock withpage,total,has_more,next_page. - Progress reporting via MCP Context — tools that make multiple API calls (
allure_get_project_statistics,allure_search_failed_tests) andallure_list_test_casesemitctx.report_progress+ctx.infoevents so compatible clients can render progress bars and step labels.
Features
6 tools covering everyday Allure TestOps workflows:
Discovery
allure_list_projects— all projects with ID, name, abbreviationallure_get_project_statistics— TC count, automation rate, last launch summary
Launches & results
allure_list_launches— recent launches with pass/fail statsallure_get_test_results— test results in a launch (filter by status)allure_search_failed_tests— FAILED/BROKEN tests in last or specified launch
Test cases
allure_list_test_cases— test cases with automated/manual filter (each result also carries its layer, e.g.UNIT/API/E2E)
Installation
Requires Python 3.10+.
# via uvx (recommended)
uvx --from allure-testops-mcp allure-testops-mcp
# or via pipx
pipx install allure-testops-mcp
Configuration
Short version — claude mcp add:
claude mcp add allure -s project \
--env ALLURE_URL=https://allure.example.com \
--env ALLURE_TOKEN=your-api-token \
--env ALLURE_SSL_VERIFY=true \
-- uvx --from allure-testops-mcp allure-testops-mcp
Or in ~/.claude.json / project .mcp.json:
{
"mcpServers": {
"allure": {
"type": "stdio",
"command": "uvx",
"args": ["--from", "allure-testops-mcp", "allure-testops-mcp"],
"env": {
"ALLURE_URL": "https://allure.example.com",
"ALLURE_TOKEN": "${ALLURE_TOKEN}",
"ALLURE_SSL_VERIFY": "true"
}
}
}
}
See .env.example for a template of all supported environment variables.
Check:
claude mcp list
# allure: uvx --from allure-testops-mcp allure-testops-mcp - ✓ Connected
Environment variables
| Variable | Required | Description |
|---|---|---|
ALLURE_URL |
yes | Allure TestOps URL (e.g. https://allure.example.com) |
ALLURE_TOKEN |
yes | API token from Allure TestOps (Profile → API tokens) |
ALLURE_SSL_VERIFY |
no | true/false. Set to false for self-signed corp certs. Default: true. |
Example usage
In Claude Code:
- "List all Allure projects"
- "Show last 10 launches for project 63"
- "Failed tests in the last launch for project 175"
- "Automation rate for project 842"
- "Test results in launch 12345 with status FAILED"
Security considerations
- API token is read from
ALLURE_TOKENenv var — never passed on the command line and never written to logs. - Secrets are not echoed back in tool responses (no
stat.request_headersdumps, nosession.authreflection). - Self-signed SSL is opt-in via
ALLURE_SSL_VERIFY=false— the default istrue. Disabling verification on a public network is a security risk; only use for trusted corporate instances. - Proxy discovery is disabled (
session.trust_env = False) — the MCP deliberately ignoresHTTP_PROXY/HTTPS_PROXYenv vars so the session cannot be silently routed through an unintended proxy. If your Allure instance is reachable only via proxy, run the MCP in an environment whererequestscan resolve directly. - No write operations exposed — all 6 tools are read-only. Even if the API token has write scope, this MCP server cannot create, modify, or delete anything in Allure TestOps.
- Input validation via Pydantic — every tool argument is typed and bounded (IDs must be ≥ 1, pagination capped at 200-500).
Rate limits
Allure TestOps enforces per-instance rate limits (typically ~60 requests / minute for API tokens). On HTTP 429 the MCP returns an actionable error suggesting you:
- Wait 30-60 seconds before retrying.
- Reduce the
sizeparameter (default 50 for test results, 200 for projects). - Paginate with smaller page sizes.
Two tools perform multiple API calls internally:
allure_get_project_statistics— 3 calls (TC counts + launches + launch statistic).allure_search_failed_tests— 2-3 calls (latest launch resolve + FAILED + BROKEN).
Both use MCP Context to report per-step progress; allure_list_test_cases also emits a single progress event. Monitor the progress stream in compatible clients.
Development
git clone https://github.com/mshegolev/allure-testops-mcp.git
cd allure-testops-mcp
pip install -e '.[dev]'
pytest
Run the server directly (stdio transport, waits on stdin for MCP messages):
ALLURE_URL=... ALLURE_TOKEN=... allure-testops-mcp
License
MIT © Mikhail Shchegolev
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。