← 返回 Skills 市场
yedanyagamiai-cmd

OpenClaw AgentForge Compare

作者 YedanYagami · GitHub ↗ · v2.0.0
cross-platform ✓ 安全检测通过
283
总下载
0
收藏
1
当前安装
2
版本数
在 OpenClaw 中安装
/install agentforge-compare
功能描述
AI coding tool comparison MCP server -- compare Claude Code vs Cursor vs Windsurf vs Devin vs Copilot side-by-side with feature matrices, pricing breakdowns,...
使用说明 (SKILL.md)

\r \r

OpenClaw AgentForge Compare v2.0\r

\r Pick the right AI tool -- 5 tools to compare, profile, recommend, and calculate costs for 15+ AI coding assistants across 7 dimensions.\r \r

Quick Reference\r

\r | You say... | Protocol | Tool | Tier |\r |------------|----------|------|------|\r | "Compare Cursor vs Windsurf" | BattleCard | compare_tools | Free |\r | "Tell me about Devin" | DeepProfile | get_tool_profile | Free |\r | "List AI coding tools" | Arsenal | list_categories | Free |\r | "Which tool for my team?" | MatchMaker | get_recommendations | Pro |\r | "How much would Cursor cost us?" | CostRadar | pricing_calculator | Pro |\r \r

Quick Start\r

\r

{\r
  "openclaw-agentforge": {\r
    "type": "streamable-http",\r
    "url": "https://agentforge-compare-mcp.yagami8095.workers.dev/mcp"\r
  }\r
}\r
```\r
\r
Add to Claude Desktop, Cursor, Windsurf, VS Code, or any MCP-compatible client. Works immediately -- no API key required for Free tier.\r
\r
## What's New in v2.0\r
\r
- :crossed_swords: **15+ Tools Tracked** -- Claude Code, Cursor, Windsurf, Devin, GitHub Copilot, Cody, Aider, Continue, Tabnine, Amazon Q, Codeium, JetBrains AI, Replit Agent, and more. Updated weekly.\r
- :brain: **AI-Powered Recommendations** (Pro) -- Describe your use case, team size, budget, and must-haves. Get ranked recommendations with match scores and migration tips.\r
- :abacus: **TCO Calculator** (Pro) -- Total cost of ownership factoring team seats, usage patterns, add-ons, and annual vs. monthly billing. Includes per-developer breakdown.\r
- :bar_chart: **7 Comparison Dimensions** -- Code generation, debugging, refactoring, chat, terminal, multi-file editing, and agent mode.\r
\r
## Detection Triggers\r
\r
This skill activates when the user:\r
\r
| Trigger Phrase | Maps To |\r
|----------------|---------|\r
| "compare Claude Code and Cursor" / "Cursor vs Windsurf" | `compare_tools` |\r
| "which AI coding tool" / "best AI assistant" | `compare_tools` |\r
| "Devin vs Copilot" / "compare AI tools" | `compare_tools` |\r
| "tell me about Cursor" / "Windsurf features" | `get_tool_profile` |\r
| "what is Devin" / "Copilot pricing" | `get_tool_profile` |\r
| "list AI coding tools" / "browse by category" | `list_categories` |\r
| "what tools exist" / "AI tool categories" | `list_categories` |\r
| "recommend an AI tool for" / "which tool for my team" | `get_recommendations` |\r
| "best tool for startups" / "best for solo dev" | `get_recommendations` |\r
| "how much does Cursor cost" / "calculate AI tool costs" | `pricing_calculator` |\r
| "AI tool pricing" / "annual vs monthly" | `pricing_calculator` |\r
\r
## Named Protocols\r
\r
### BattleCard Protocol -- Side-by-Side Comparison Engine\r
Generates structured comparison matrices for 2-5 AI coding tools across 7 dimensions: code generation quality, debugging capabilities, refactoring power, chat interface, terminal integration, multi-file editing, and autonomous agent mode. Each dimension is rated 1-10 with a summary verdict.\r
\r
### DeepProfile Protocol -- Comprehensive Tool Dossier\r
Builds a complete profile for any single tool: description, all pricing tiers (free/pro/enterprise), supported AI models, IDE integrations, key features, strengths, weaknesses, best-for scenarios, alternatives, and recent changelog highlights. Think of it as the "Wirecutter review" for AI tools.\r
\r
### Arsenal Protocol -- Category Browser\r
Organizes all 15+ tracked tools into categories: IDE-integrated assistants, CLI-based agents, cloud agents, code review tools, and specialized tools (testing, docs, security). Each category includes tool count, top picks, and quick comparison notes.\r
\r
### MatchMaker Protocol -- AI Recommendation Engine (Pro)\r
Takes your use case description, team size, monthly budget, must-have features, and preferred IDE, then returns top 3 tools ranked by match score with detailed reasoning, tradeoffs, and migration tips from your current setup.\r
\r
### CostRadar Protocol -- Total Cost Calculator (Pro)\r
Computes monthly and annual costs for any tool configuration: team seats, plan tier, estimated usage (requests/day), and add-ons. Returns per-developer cost and automatically compares against alternatives at the same scale.\r
\r
## Tools (5)\r
\r
### `compare_tools` -- BattleCard Protocol\r
Compare 2-5 AI coding tools side-by-side across all dimensions.\r
\r
**Input:**\r
- `tools` -- Array of tool names, e.g., `["Claude Code", "Cursor", "Windsurf"]`\r
\r
**Output:**\r
- Feature comparison matrix (7 dimensions, rated 1-10)\r
- Pricing breakdown per tool\r
- Model support comparison\r
- IDE compatibility matrix\r
- Strengths/weaknesses per tool\r
- Verdict summary with winner per dimension\r
\r
**Tracked Tools:** Claude Code, Cursor, Windsurf, Devin, GitHub Copilot, Sourcegraph Cody, Aider, Continue, Tabnine, Amazon Q Developer, Codeium, JetBrains AI Assistant, Replit Agent, Sweep, and more.\r
\r
### `get_tool_profile` -- DeepProfile Protocol\r
Get a comprehensive profile for any single AI coding tool.\r
\r
**Input:**\r
- `tool_name` -- Name of the tool (e.g., `"Cursor"`)\r
\r
**Output:**\r
- Description and positioning\r
- Pricing: free tier, pro tier, enterprise tier with exact prices\r
- Supported AI models (GPT-4, Claude, etc.)\r
- IDE integrations\r
- Key features list\r
- Strengths and weaknesses\r
- Best-for scenarios\r
- Top alternatives\r
- Recent changelog highlights\r
\r
### `list_categories` -- Arsenal Protocol\r
Browse all tracked AI tools organized by category.\r
\r
**Output:**\r
- **IDE-Integrated**: Cursor, Windsurf, Copilot, Cody, Tabnine, Codeium, JetBrains AI, Amazon Q\r
- **CLI-Based**: Claude Code, Aider, Continue\r
- **Cloud Agents**: Devin, Replit Agent, Sweep\r
- **Code Review**: Copilot PR Review, CodeRabbit, Sourcery\r
- **Specialized**: Testing, documentation, security-focused tools\r
\r
Each category: tool count, top picks, brief comparison notes.\r
\r
### `get_recommendations` (Pro) -- MatchMaker Protocol\r
Describe your situation and get AI-powered ranked recommendations.\r
\r
**Input:**\r
- `use_case` -- What you're building or doing (e.g., "full-stack web app, React + Node")\r
- `team_size` -- Number of developers\r
- `budget` -- Monthly budget per developer\r
- `must_have` -- Required features (e.g., "multi-file editing, agent mode")\r
- `preferred_ide` -- VS Code, JetBrains, Neovim, etc.\r
\r
**Output:**\r
- Top 3 tools ranked by match score (0-100)\r
- Detailed reasoning for each recommendation\r
- Tradeoff analysis (what you gain vs. what you lose)\r
- Migration tips from your current tool\r
\r
### `pricing_calculator` (Pro) -- CostRadar Protocol\r
Calculate complete costs for any AI tool configuration.\r
\r
**Input:**\r
- `tool_name` -- Which tool to calculate\r
- `team_size` -- Number of seats\r
- `plan` -- `free`, `pro`, or `enterprise`\r
- `usage` -- Estimated requests per day per developer\r
- `add_ons` -- Any additional features needed\r
\r
**Output:**\r
- Monthly cost (total and per-developer)\r
- Annual cost with savings vs. monthly\r
- Feature breakdown at selected tier\r
- Automatic comparison with 2 alternatives at same scale and budget\r
\r
## Wrong vs. Right\r
\r
### Example 1: Too Many Tools in One Comparison\r
\r
**Wrong:**\r
```\r
User: "Compare all AI coding tools"\r
Agent: [calls compare_tools with 12 tools]  // Unwieldy, unreadable output\r
```\r
\r
**Right:**\r
```\r
User: "Compare all AI coding tools"\r
Agent: "There are 15+ tools tracked. Let me help narrow it down. Are you looking for:\r
  - IDE-integrated (Cursor, Windsurf, Copilot)?\r
  - CLI-based (Claude Code, Aider)?\r
  - Cloud agents (Devin, Replit)?\r
Pick a category and I'll compare the top 2-3."\r
```\r
\r
### Example 2: Stale Pricing Assumptions\r
\r
**Wrong:**\r
```\r
User: "How much is Cursor?"\r
Agent: "Cursor Pro is $20/mo."  // Recalled from training data, may be outdated\r
```\r
\r
**Right:**\r
```\r
User: "How much is Cursor?"\r
Agent: [calls get_tool_profile("Cursor")]  // Gets latest tracked pricing\r
Agent: "According to our latest data (updated weekly), Cursor Pro is $20/mo. Always verify on cursor.com for flash sales or plan changes."\r
```\r
\r
### Example 3: Recommendation Without Context\r
\r
**Wrong:**\r
```\r
User: "Which AI tool should I use?"\r
Agent: [calls get_recommendations with empty fields]  // Garbage in, garbage out\r
```\r
\r
**Right:**\r
```\r
User: "Which AI tool should I use?"\r
Agent: "To give you a good recommendation, I need a few details:\r
  1. What are you building? (web app, mobile, data science, etc.)\r
  2. Team size? (solo, small team, enterprise)\r
  3. Monthly budget per developer?\r
  4. Must-have features? (agent mode, multi-file editing, etc.)\r
  5. Preferred IDE? (VS Code, JetBrains, Neovim, etc.)"\r
```\r
\r
## Security & Privacy\r
\r
| Concern | How We Handle It |\r
|---------|------------------|\r
| **No Tool Telemetry** | We don't track which tools you compare or recommend. |\r
| **No Team Data Stored** | Team size, budget, and use case inputs are processed and discarded -- never logged. |\r
| **Pricing Source** | All pricing data sourced from official public pages. No scraping of private accounts. |\r
| **Rate Limiting** | Per-IP limits prevent abuse. Free: 10/day, Pro: 1000/day. |\r
| **Edge Processing** | All computation on Cloudflare Workers -- no centralized database of user queries. |\r
| **Unbiased Results** | No affiliate deals or sponsored rankings. Comparisons are data-driven. |\r
\r
## Pricing\r
\r
| Tier | Calls/Day | Price | What You Get |\r
|------|-----------|-------|--------------|\r
| **Free** | 10 | $0/mo | 3 tools: compare, profile, categories |\r
| **Pro** | 1,000 | $9/mo | All 5 tools + recommendations + calculator + all 9 OpenClaw servers |\r
| **x402** | Pay-per-call | $0.05 USDC | No account needed, crypto-native |\r
\r
**Get Pro Key**: https://buy.stripe.com/4gw5na5U19SP9TW288\r
\r
One Pro key unlocks all 9 servers. Cancel anytime.\r
\r
## The OpenClaw Intelligence Stack\r
\r
| Server | Tools | Best For |\r
|--------|-------|----------|\r
| **JSON Toolkit** | 6 | Format, validate, diff, query, transform JSON |\r
| **Regex Engine** | 5 | Test, extract, replace, explain regex patterns |\r
| **Color Palette** | 5 | Generate, convert, harmonize, accessibility-check colors |\r
| **Timestamp Converter** | 5 | Parse, format, diff, timezone-convert timestamps |\r
| **Prompt Enhancer** | 6 | Optimize, rewrite, score, A/B test AI prompts |\r
| **Market Intelligence** | 6 | AI market trends, GitHub stats, competitor analysis |\r
| **Fortune & Tarot** | 3 | Daily fortune, tarot readings, I Ching |\r
| **Content Publisher** | 8 | MoltBook posts, social content, newsletter |\r
| **AgentForge Compare** | **5** | **Compare AI tools, frameworks, MCP servers** |\r
\r
All 9 servers share one Pro key. **$9/mo = 49 tools.**\r
安全使用建议
This skill appears to do what it says (compare AI coding tools) and requires no local installs or credentials. Important precautions before installing or using it: (1) the skill forwards your prompts and any context to a third‑party Cloudflare Workers endpoint (agentforge-compare-mcp.yagami8095.workers.dev). Do not send proprietary code, secrets, or sensitive configuration through it until you trust the operator. (2) Verify the skill's source: check the referenced GitHub repo, confirm the maintainer identity, and prefer a published repository or official homepage if available. (3) If you need comparisons on sensitive internal tools, ask for a self-hostable or on-premises option, or request the skill author to publish source and a privacy/security statement. (4) Be cautious with the advertised 'Pro' billing—confirm payment flow and data retention policies before subscribing.
功能分析
Type: OpenClaw Skill Name: agentforge-compare Version: 2.0.0 The agentforge-compare skill bundle is a functional tool designed to provide AI coding assistant comparisons and pricing information via a remote MCP server. The SKILL.md file contains well-structured instructions and tool definitions (such as compare_tools and get_tool_profile) that align with its stated purpose, while the _meta.json file provides standard metadata. No evidence of data exfiltration, malicious code execution, or harmful prompt injection was found; the use of a remote endpoint (agentforge-compare-mcp.yagami8095.workers.dev) is consistent with the described streamable-http architecture.
能力评估
Purpose & Capability
Name/description (AI coding tool comparisons, pricing, recommendations) align with the instructions and declared protocols (BattleCard, DeepProfile, CostRadar). There are no unrelated env vars, binaries, or installs requested.
Instruction Scope
The SKILL.md instructs the agent to act as an MCP client for a remote streamable-http MCP at https://agentforge-compare-mcp.yagami8095.workers.dev/mcp. It does not instruct the agent to read local files or secrets, but as an external MCP it will receive full user prompts and any context the agent sends — so user data (including code, proprietary details, or secrets) will be forwarded to that third‑party endpoint.
Install Mechanism
Instruction-only skill with no install spec and no code files; nothing is written to disk by the skill itself.
Credentials
The skill requests no environment variables or credentials, which is proportionate to its stated purpose. However, it relies on a third‑party HTTP endpoint (a personal Cloudflare Workers domain) that will receive user prompts — this raises a data-exposure concern even though no credentials are requested.
Persistence & Privilege
always is false and the skill does not request elevated platform privileges or modify other skills. Autonomous invocation is allowed by default (not flagged by itself).
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install agentforge-compare
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /agentforge-compare 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v2.0.0
v2.0.0 — Named Protocols (BattleCard, DeepProfile, Arsenal, MatchMaker, CostRadar), trigger-rich descriptions, Wrong/Right examples
v1.0.0
- Initial release of openclaw-agentforge-compare (v1.0.0) - Compare popular AI coding assistants (Claude Code, Cursor, Windsurf, Devin, Copilot, and more) side-by-side - Access detailed tool profiles including features, pricing, pros, and cons - Get comprehensive pricing tables for all supported tools - AI-powered tool recommendations available with Pro key - Hosted on Cloudflare Workers for zero-install, remote usage
元数据
Slug agentforge-compare
版本 2.0.0
许可证
累计安装 1
当前安装数 1
历史版本数 2
常见问题

OpenClaw AgentForge Compare 是什么?

AI coding tool comparison MCP server -- compare Claude Code vs Cursor vs Windsurf vs Devin vs Copilot side-by-side with feature matrices, pricing breakdowns,... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 283 次。

如何安装 OpenClaw AgentForge Compare?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install agentforge-compare」即可一键安装,无需额外配置。

OpenClaw AgentForge Compare 是免费的吗?

是的,OpenClaw AgentForge Compare 完全免费(开源免费),可自由下载、安装和使用。

OpenClaw AgentForge Compare 支持哪些平台?

OpenClaw AgentForge Compare 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 OpenClaw AgentForge Compare?

由 YedanYagami(@yedanyagamiai-cmd)开发并维护,当前版本 v2.0.0。

💬 留言讨论