About Tencent Cloud Hackathon Intelligent Pentest Competition API Server
Tencent Cloud Hackathon Intelligent Pentest Competition API Server is an MCP server published by Neuro-Sploit in the Developer Tools category: mCP server: Tencent Cloud Hackathon Intelligent Pentest Competition API Server.
The repository has 45 stars and 11 forks, with the last commit 10 months ago. Six months or more without a commit doesn't mean the server is broken, but check the open issues before depending on it in production.
Install
git clone https://github.com/Neuro-Sploit/tencent-cloud-hackathon-intelligent-pentest-competition-api-serverThis server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.
Ask AI
Ask AI about Tencent Cloud Hackathon Intelligent Pentest Competition API Server
Powered by Claude · Grounded in docs
Security checks
Static checks derived from the repository 7 days ago. Heuristics on the source, not an audit.
- ✓README presentThe repository ships documentation.
- !License declaredNo license file. Legally you have no right to use it in a product.
- !Tests presentNo tests found. Behaviour changes may go unnoticed between versions.
- ✓Dependencies pinnedA lockfile pins dependency versions.
- ✓No dynamic code executionSource scan found no eval, dynamic Function, or shell execution.
- !Scoped permissionsDoesn't declare a permission scope. Assume it can do anything its process can.
Releases
README
Tencent Cloud Intelligent Pentest Competition Mock API Server
本项目为腾讯云智能化渗透测试大赛 API 模拟器,旨在本地复现赛题下发、提示获取及答题流程,辅助自动化攻防策略的开发与调试。
功能特性
- 靶场编排:基于官方 XBow benchmark 动态复制赛题环境,自动映射端口并通过 Docker Compose 启动容器。
- 协议仿真:完全复刻线上平台 API 结构,覆盖题目获取、提示查询、答案提交等核心接口。
- 多端支持:提供 CLI、Python SDK 及 MCP (FastMCP) 协议支持,便于脚本调用及 AI Agent 集成。
- 资源管理:全链路结构化日志记录;进程退出时自动清理靶场容器,防止资源泄露。
环境要求
- Python 环境:建议使用 uv 管理依赖。
- Container Runtime:安装 Docker 及 Docker Compose。
- Redis:本地需启动 Redis 服务以模拟官方 API 的速率限制 (1 QPS)。
- Benchmark 数据:需下载修正版题目环境。
- 官方环境存在缺陷,请使用修正版:Neuro-Sploit/xbow-validation-benchmarks。
部署与启动
1. 准备赛题数据
git clone https://github.com/Neuro-Sploit/xbow-validation-benchmarks --branch main --depth 1 ~/xbow-validation-benchmarks
2. 启动模拟服务器

# 克隆仓库
git clone https://github.com/WangYihang/tencent-cloud-hackathon-intelligent-pentest-competition-api-server.git
cd tencent-cloud-hackathon-intelligent-pentest-competition-api-server
# 安装依赖
uv sync
# 启动服务
# --host: API 监听地址
# --port: API 监听端口
# --public-accessible-host: 赛题对外暴露的 IP (通常为本机局域网 IP,如:192.168.1.2)
# -i: 指定启动的题目 ID (如 XBEN-001-24),该选项可以通过重复指定来同时启动多个题目环境
python -m tencent_cloud_hackathon_intelligent_pentest_competition_api_server.server \
--xbow-benchmark-folder ~/xbow-validation-benchmarks/benchmarks \
--host 0.0.0.0 \
--port 8000 \
--public-accessible-host 192.168.1.2 \
-i 1 -i 2 -i 3 -i 4
服务启动后相关资源地址:
- Swagger UI: http://127.0.0.1:8000/docs
- 运行日志: logs/competition-platform-server-logs.jsonl

API 接口
接口协议遵循官方文档。
| 方法 | 路径 | 描述 |
|---|---|---|
| GET | /api/v1/challenges | 获取当前阶段题目实例信息(目标 IP、端口、积分等)。 |
| GET | /api/v1/hint/{challenge_code} | 获取题目提示(首次查看将触发罚分机制)。 |
| POST | /api/v1/answer | 提交 Flag,校验通过后返回积分并标记完成。 |
客户端
客户端工具默认读取以下环境变量:
export COMPETITION_BASE_URL=http://127.0.0.1:8000
export COMPETITION_API_TOKEN=00000000-0000-0000-0000-000000000000
命令行工具 (CLI)

# 获取题目列表
python -m tencent_cloud_hackathon_intelligent_pentest_competition_api_server.client_cli get-challenges
# 获取提示
python -m tencent_cloud_hackathon_intelligent_pentest_competition_api_server.client_cli get-challenge-hint <challenge_code>
# 提交答案
python -m tencent_cloud_hackathon_intelligent_pentest_competition_api_server.client_cli submit-answer <challenge_code> <flag>
Python SDK
- Python SDK: client_sdk.APIClient 内置指数退避重试策略与速率限制处理。
pip install tencent-cloud-hackathon-intelligent-pentest-competition-api-server
from tencent_cloud_hackathon_intelligent_pentest_competition_api_server.client_sdk import APIClient
client = APIClient(
base_url='http://127.0.0.1:8000',
api_token='00000000-0000-0000-0000-000000000000',
)
challenges = client.get_challenges()
print(challenges)
first_challenge_code = challenges.challenges[0].challenge_code
print(client.get_challenge_hint(first_challenge_code))
print(client.submit_answer(first_challenge_code, 'flag{...}'))
MCP Server
- MCP Protocol: 通过 client_mcp 模块暴露接口,支持 AI Agent 直接调用。
> python -m tencent_cloud_hackathon_intelligent_pentest_competition_api_server.client_mcp
{"timestamp": "2025-11-27T15:57:31.881942", "level": "INFO", "message": "Initializing API client", "action": "init", "base_url": "http://localhost:8000/", "api_token": "00000000-0000-0000-0000-000000000000"}
╭──────────────────────────────────────────────────────────────────────────────╮
│ │
│ ▄▀▀ ▄▀█ █▀▀ ▀█▀ █▀▄▀█ █▀▀ █▀█ │
│ █▀ █▀█ ▄▄█ █ █ ▀ █ █▄▄ █▀▀ │
│ │
│ FastMCP 2.13.1 │
│ │
│ │
│ 🖥 Server name: Capture The Flag Competition API │
│ │
│ 📦 Transport: STDIO │
│ │
│ 📚 Docs: https://gofastmcp.com │
│ 🚀 Hosting: https://fastmcp.cloud │
│ │
╰──────────────────────────────────────────────────────────────────────────────╯
[11/27/25 15:57:31] INFO Starting MCP server 'Capture The Flag Competition API' with transport 'stdio'
运维说明
- 生命周期管理:ChallengeManager 在启动时分配随机端口,进程终止 (SIGINT/SIGTERM) 时自动执行 docker compose down 并清理临时文件。
- 日志分析:服务与客户端行为均记录于 logs/ 目录下的 JSONL 文件。
- 故障排查:
- 端口冲突:若自动分配端口失败,请删除 challenges/ 目录下的临时文件夹后重启。
- Flag 读取异常:确保 Benchmark 环境 .env 文件中 FLAG 变量已正确设置。
README fetched from the repository 7 days ago. The original is authoritative.