weapp-agent
Documentation and tooling are built using Cloudflare's Nimbus framework.
Supports OpenAI Responses as a model provider, enabling the agent to run tasks using OpenAI models.
Targets weapp-vite native projects, providing delivery capabilities such as engineering diagnostics, deterministic acceptance, and screenshots/logs.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@weapp-agentrun acceptance checks on this Weapp project and show me the report"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Weapp Agent
让 AI 写完的小程序代码,更容易被验证和交付。
面向 weapp-vite 原生与 Wevu 微信项目的交付能力层。通过 MCP 和 CLI 为 Codex 等宿主提供工程诊断、确定性验收与截图日志,不需要第二个模型或 API Key。保留独立 AI CLI 模式。
当前状态
源码预览版。仓库提供可运行实现和可打包 CLI。实际验收记录见 VALIDATION.md,模拟测试与真实模型、DevTools 验证分别记录。
Related MCP server: wechat-devtools-mcp
安装预览包
从 npm 安装预览版:
npm install --global @weapp-agent/cli@preview
weapp-agent --version从源码开始
需要 Node.js 24.15+、pnpm。
corepack enable
pnpm install
pnpm build
node apps/cli/dist/index.mjs --help
pnpm --filter @weapp-agent/cli pack --pack-destination ../../artifacts将生成的 .tgz 安装到全局,或直接通过 node 调用构建入口。
使用已有 AI 宿主时,先阅读无模型接入指南。以下命令用于独立 agent 模式:
weapp-agent init --provider openai --model YOUR_MODEL
# 在终端环境中设置 OPENAI_API_KEY
weapp-agent doctor
weapp-agent --trust run "为首页增加一个计数器,执行构建与测试"
weapp-agent新建项目:
weapp-agent init my-miniapp --create --template wevu --model YOUR_MODEL
cd my-miniapp
pnpm install
weapp-agent --trust在现有 AI 工具中验收
以下新入口需要包含本次改动的构建包,尚未发布到 npm:
weapp-agent init
weapp-agent doctor --json
# 审阅项目脚本、配置与场景后
weapp-agent --trust accept --json
weapp-agent -C /absolute/path/to/project mcp
weapp-agent skill .agents/skills/weapp-acceptance在配置的 acceptance.scenarios 中列出场景 JSON 文件。默认要求 build 和 devtools 都通过;缺少场景返回未验证。MCP 提供 inspect/start/status/cancel/report 五类操作,宿主负责代码修改和修复。新报告为 version 2,绑定源码摘要,支持取消和中断识别,报告不会自动重放交互。旧 verify 输出继续兼容。
接入、场景格式与排错 · 计数器场景 · 实施状态与发布门槛 · 对照评测
工作方式
模型独立: OpenAI Responses、Anthropic Messages、OpenAI-compatible;支持参考截图。
修改可检查: 流式进度、代码差异、文件哈希冲突检测,保留用户现有改动。
验证有边界: 类型检查、构建、测试、DevTools 分别报告,缺失检查标记未验证。
会话可恢复: 本地 JSONL 日志,不自动重放中断的副作用操作。
权限可理解: 可信项目内编辑和已配置检查自动执行;任意 Shell、未知 MCP、上传发布需确认。
自动化接口:
run --json输出版本化事件;verify --json输出检查报告。
weapp-agent run "参考截图实现首页" --image reference.png
weapp-agent sessions
weapp-agent resume SESSION_ID "继续修复失败的检查"
weapp-agent verify --json--trust 表示你已审阅该项目脚本与配置。工具权限控制不是 OS 沙箱。任务内容和所需源码、截图会发往配置的模型服务,会话保存在本机。不要在提示词或配置中放置密钥。
开发
pnpm validate
pnpm exec repo doctor
pnpm exec repo check
pnpm docs:dev工程由 repoctl 创建,文档由 Nimbus 创建。核心、模型适配、小程序工具和终端界面相互分离。设计参考与固定源码版本见 架构文档。
License
MIT。第三方项目的商标与许可证归各自维护者所有。
This server cannot be deployed
Maintenance
Related MCP Connectors
Build, validate, deploy — HTTP APIs, cron jobs, webhooks and MCP tools — from your AI client.
Deploy and manage Jade Hosting projects from AI clients. Jade account and OAuth required.
AI-native git hosting — repos, PRs, issues, CI gates, and AI code review over MCP (60 tools).
Pre-execution governance for AI agents. Deterministic PASS/FAIL/REVIEW verdicts, replayable proof.
Related MCP Servers
- AlicenseBqualityDmaintenanceEnables automated testing of WeChat mini-programs via Model Context Protocol, providing tools for connecting to WeChat Developer Tools, querying and interacting with page elements, making assertions, navigating, and debugging.2024 npm78MIT
- AlicenseAqualityAmaintenanceWraps the WeChat DevTools CLI as an MCP service, enabling AI in editors to call WeChat CLI commands for mini-program development, testing, debugging, and automation.7210 PyPI150MIT
- -licenseNot gradedqualityNot gradedmaintenanceEnables AI agents to debug and control WeChat Mini Programs through WeChat DevTools, including direct automator operations and CLI-backed features like screenshots, logs, previews, and uploads.-
- AlicenseNot gradedqualityBmaintenanceEnables evaluating AI applications, inspecting reliability evidence, and gating releases from development and CI workflows.32Apache 2.0