/ 目录 / 演练场 / convex-agent-skillz
● 社区 PolarCoding85 ⚡ 即开即用

convex-agent-skillz

作者 PolarCoding85 · PolarCoding85/convex-agent-skillz

Drop-in Convex backend knowledge — queries, mutations, real-time presence, workflows, auth (Clerk/WorkOS), and AI agent patterns auto-load by context.

PolarCoding85/convex-agent-skillz gives Claude deep expertise across Convex: queries/mutations/actions/HTTP endpoints, schema + indexing, real-time features (presence, streaming, collab editing), durable operations (workflows, queues, schedules), auth (universal + Clerk + WorkOS), AI integration (agent threads, tool calling, RAG, message history). Context-aware — working in convex/ auto-activates the convex-expert agent; referencing components like Rate Limiter or Stripe loads the matching pattern.

为什么要用

核心特性

实时演示

实际使用效果

convex-agent-skillz-skill.replay ▶ 就绪
0/0

安装

选择你的客户端

~/Library/Application Support/Claude/claude_desktop_config.json  · Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "convex-agent-skillz-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/PolarCoding85/convex-agent-skillz",
        "~/.claude/skills/convex-agent-skillz"
      ],
      "_inferred": true
    }
  }
}

打开 Claude Desktop → Settings → Developer → Edit Config。保存后重启应用。

~/.cursor/mcp.json · .cursor/mcp.json
{
  "mcpServers": {
    "convex-agent-skillz-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/PolarCoding85/convex-agent-skillz",
        "~/.claude/skills/convex-agent-skillz"
      ],
      "_inferred": true
    }
  }
}

Cursor 使用与 Claude Desktop 相同的 mcpServers 格式。项目级配置优先于全局。

VS Code → Cline → MCP Servers → Edit
{
  "mcpServers": {
    "convex-agent-skillz-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/PolarCoding85/convex-agent-skillz",
        "~/.claude/skills/convex-agent-skillz"
      ],
      "_inferred": true
    }
  }
}

点击 Cline 侧栏中的 MCP Servers 图标,然后选 "Edit Configuration"。

~/.codeium/windsurf/mcp_config.json
{
  "mcpServers": {
    "convex-agent-skillz-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/PolarCoding85/convex-agent-skillz",
        "~/.claude/skills/convex-agent-skillz"
      ],
      "_inferred": true
    }
  }
}

格式与 Claude Desktop 相同。重启 Windsurf 生效。

~/.continue/config.json
{
  "mcpServers": [
    {
      "name": "convex-agent-skillz-skill",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/PolarCoding85/convex-agent-skillz",
        "~/.claude/skills/convex-agent-skillz"
      ]
    }
  ]
}

Continue 使用服务器对象数组,而非映射。

~/.config/zed/settings.json
{
  "context_servers": {
    "convex-agent-skillz-skill": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/PolarCoding85/convex-agent-skillz",
          "~/.claude/skills/convex-agent-skillz"
        ]
      }
    }
  }
}

加入 context_servers。Zed 保存后热重载。

claude mcp add convex-agent-skillz-skill -- git clone https://github.com/PolarCoding85/convex-agent-skillz ~/.claude/skills/convex-agent-skillz

一行命令搞定。用 claude mcp list 验证,claude mcp remove 卸载。

使用场景

实战用法: convex-agent-skillz

Add real-time presence to an existing Convex app

👤 Convex devs adding collab features ⏱ ~45 min intermediate

何时使用: You have an app and want Figma-style 'who's here' indicators without running a socket server.

前置条件
  • Skill copied into your project's .claude folder — git clone https://github.com/PolarCoding85/convex-agent-skillz and copy .claude/ to your project root
步骤
  1. Describe the feature
    Using the Convex skillz — add presence to the documents page. Show cursors and usernames.✓ 已复制
    → Claude implements with Convex's presence pattern (mutation + reactive query), not a custom WebSocket
  2. Handle disconnection
    What about cleanup when users close the tab?✓ 已复制
    → Specific stale-presence cleanup strategy aligned with Convex conventions

结果: Working presence with proper cleanup.

注意事项
  • Custom presence pattern that fights Convex's reactivity — Skill steers you toward the idiomatic version; don't override without a reason
搭配使用: filesystem

Integrate Clerk auth with Convex properly

👤 Devs wiring Clerk into a new or existing Convex app ⏱ ~30 min intermediate

何时使用: You picked Clerk and want its Convex integration right the first time.

步骤
  1. Wire the provider
    Convex skillz — integrate Clerk auth with Convex. Set up the JWT template and the Convex auth config.✓ 已复制
    → Claude references the right Clerk JWT template name and auth.config.ts setup
  2. Protect queries
    Now gate the documents query so only the owner can read.✓ 已复制
    → ctx.auth.getUserIdentity() pattern with proper identity check

结果: Auth that passes code review and doesn't leak across tenants.

Build an AI agent with persistent threads on Convex

👤 Devs building chat or agent products ⏱ ~90 min advanced

何时使用: You want threaded message history that survives across sessions.

步骤
  1. Model threads
    Convex skillz — agent threads table + message history + tool calling metadata.✓ 已复制
    → Schema that covers streaming, partial messages, and tool-call state
  2. Hook up streaming
    Stream the assistant response back via a reactive query.✓ 已复制
    → Partial-message pattern leveraging Convex's reactivity

结果: A Convex-native agent thread layer ready to scale.

组合

与其他 MCP 搭配,撬动十倍杠杆

convex-agent-skillz-skill + github

Open PRs for feature additions with Convex-conventional diffs

Implement the presence feature, commit and open PR with the Convex skill patterns noted.✓ 已复制
convex-agent-skillz-skill + filesystem

Work across schema, functions, and UI in one session

Update convex/schema.ts, add the mutation, and wire up the React hook.✓ 已复制

工具

此 MCP 暴露的能力

工具输入参数何时调用成本
queries_mutations_actions feature intent Any convex/ work 0
realtime_patterns feature (presence / streaming / collab) Real-time features 0
auth_setup provider (Clerk / WorkOS / custom) Wiring auth 0
durable_ops workflow intent Long-running or retry-heavy logic 0
ai_agent_threads agent feature description Agent products on Convex 0

成本与限制

运行它的成本

API 配额
None from the skill
每次调用 Token 数
Moderate — patterns load as needed by context
费用
Free — skills are local files. Convex itself has its own pricing.
提示
Place skill in project .claude/ rather than global — auto-activation works best scoped to the project.

安全

权限、密钥、影响范围

凭据存储: No credentials in the skill. Your auth provider secrets live in your own env.
数据出站: None from the skill — code stays local

故障排查

常见错误与修复

Skill not invoked — Claude writes non-Convex-idiomatic code

Say 'use the Convex skillz' or ensure you're working inside convex/ directory.

验证: ls <project>/.claude/agents/convex-expert.md
SKILL.md frontmatter wrong

Re-copy .claude/ from the repo.

验证: head <project>/.claude/skills/*/SKILL.md
Wrong trigger keywords — auth pattern doesn't use your provider

Name the provider explicitly ('with Clerk' / 'with WorkOS').

替代方案

convex-agent-skillz 对比其他方案

替代方案何时用它替代权衡
Convex official docsYou want primary source rather than a curated skillNo in-agent auto-activation
Hand-rolled patternsYou're an experienced Convex devYou'll reinvent patterns; skill just makes Claude use yours consistently

更多

资源

📖 阅读 GitHub 上的官方 README

🐙 查看未解决的 issue

🔍 浏览全部 400+ MCP 服务器和 Skills