{
  "markdown": "# Polydev - Multi-Model AI Perspectives\n\n**Get unstuck faster.** Query GPT 5.2, Claude Opus 4.5, Gemini 3, and Grok 4.1 simultaneously — one API call, four expert opinions.\n\n[![npm version](https://img.shields.io/npm/v/polydev-ai.svg)](https://www.npmjs.com/package/polydev-ai)\n[![SWE-bench Verified](https://img.shields.io/badge/SWE--bench-74.6%25-brightgreen)](https://polydev.ai/articles/swe-bench-paper)\n[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)\n\n---\n\n## Why Polydev?\n\n**Stop copy-pasting between ChatGPT, Claude, and Gemini.** Get all their perspectives in your IDE with one request.\n\n| Metric | Result |\n|--------|--------|\n| **SWE-bench Verified** | 74.6% Resolve@2 |\n| **Cost vs Claude Opus** | 62% lower |\n| **Response time** | 10-40 seconds |\n\n> *\"Different models have different blind spots. Combining their perspectives eliminates yours.\"*\n\n---\n\n## Supported Models\n\n| Model | Provider | Strengths |\n|-------|----------|-----------|\n| **GPT 5.2** | OpenAI | Reasoning, code generation |\n| **Claude Opus 4.5** | Anthropic | Analysis, nuanced thinking |\n| **Gemini 3 Pro** | Google | Multimodal, large context |\n| **Grok 4.1** | xAI | Real-time knowledge, directness |\n\n---\n\n## Quick Start\n\n### 1. Get your free API token\n\n**[polydev.ai/dashboard/mcp-tokens](https://polydev.ai/dashboard/mcp-tokens)**\n\n| Tier | Messages/Month | Price |\n|------|----------------|-------|\n| **Free** | 1,000 | $0 |\n| **Pro** | 10,000 | $19/mo |\n\n### 2. Install in your IDE\n\n#### Claude Code\n\n```bash\nclaude mcp add polydev -- npx -y polydev-ai@latest\n```\n\nThen set your token:\n```bash\nexport POLYDEV_USER_TOKEN=\"pd_your_token_here\"\n```\n\nOr add to `~/.claude.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"polydev\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"polydev-ai@latest\"],\n      \"env\": {\n        \"POLYDEV_USER_TOKEN\": \"pd_your_token_here\"\n      }\n    }\n  }\n}\n```\n\n#### Cursor\n\nAdd to `~/.cursor/mcp.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"polydev\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"polydev-ai@latest\"],\n      \"env\": {\n        \"POLYDEV_USER_TOKEN\": \"pd_your_token_here\"\n      }\n    }\n  }\n}\n```\n\n#### Windsurf\n\nAdd to your MCP configuration:\n\n```json\n{\n  \"mcpServers\": {\n    \"polydev\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"polydev-ai@latest\"],\n      \"env\": {\n        \"POLYDEV_USER_TOKEN\": \"pd_your_token_here\"\n      }\n    }\n  }\n}\n```\n\n#### Cline (VS Code)\n\n1. Open Cline settings (gear icon)\n2. Go to \"MCP Servers\" → \"Configure\"\n3. Add the same JSON config as above\n\n#### OpenAI Codex CLI\n\nAdd to `~/.codex/config.toml`:\n\n```toml\n[mcp_servers.polydev]\ncommand = \"npx\"\nargs = [\"-y\", \"polydev-ai@latest\"]\n\n[mcp_servers.polydev.env]\nPOLYDEV_USER_TOKEN = \"pd_your_token_here\"\n\n[mcp_servers.polydev.timeouts]\ntool_timeout = 180\nsession_timeout = 600\n```\n\n---\n\n## Usage\n\n### Natural Language\n\nJust mention \"polydev\" or \"perspectives\" in your prompt:\n\n```\n\"Use polydev to debug this infinite loop\"\n\n\"Get perspectives on: Should I use Redis or PostgreSQL for caching?\"\n\n\"Use polydev to review this API for security issues\"\n```\n\n### MCP Tool\n\nCall the `get_perspectives` tool directly:\n\n```typescript\n{\n  \"tool\": \"get_perspectives\",\n  \"arguments\": {\n    \"prompt\": \"How should I optimize this database query?\",\n    \"user_token\": \"pd_your_token_here\"\n  }\n}\n```\n\n---\n\n## Example Response\n\n```\n🤖 Multi-Model Analysis\n\n┌─ GPT 5.2 ────────────────────────────────────────\n│ The N+1 query pattern is causing performance issues.\n│ Consider using eager loading or batch queries...\n└──────────────────────────────────────────────────\n\n┌─ Claude Opus 4.5 ────────────────────────────────\n│ Looking at the execution plan, the table scan on\n│ `users` suggests a missing index on `email`...\n└──────────────────────────────────────────────────\n\n┌─ Gemini 3 ───────────────────────────────────────\n│ The query could benefit from denormalization for\n│ this read-heavy access pattern...\n└──────────────────────────────────────────────────\n\n┌─ Grok 4.1 ───────────────────────────────────────\n│ Just add an index. The real problem is you're\n│ querying in a loop - fix that first.\n└──────────────────────────────────────────────────\n\n✅ Consensus: Add index on users.email, fix N+1 query\n💡 Recommendation: Use eager loading with proper indexing\n```\n\n---\n\n## Research\n\nOur approach achieves **74.6% on SWE-bench Verified** (Resolve@2), matching Claude Opus at 62% lower cost.\n\n| Approach | Resolution Rate | Cost/Instance |\n|----------|-----------------|---------------|\n| Claude Haiku (baseline) | 64.6% | $0.18 |\n| + Polydev consultation | 66.6% | $0.24 |\n| **Resolve@2 (best of both)** | **74.6%** | $0.37 |\n| Claude Opus (reference) | 74.4% | $0.97 |\n\n**[Read the full paper →](https://polydev.ai/articles/swe-bench-paper)**\n\n---\n\n## Available Tools\n\n| Tool | Description |\n|------|-------------|\n| `get_perspectives` | Query multiple AI models simultaneously |\n| `get_cli_status` | Check status of local CLI tools |\n| `force_cli_detection` | Re-detect installed CLI tools |\n| `send_cli_prompt` | Send prompts to local CLIs with fallback |\n\n---\n\n## Links\n\n- **Website:** [polydev.ai](https://polydev.ai)\n- **Dashboard:** [polydev.ai/dashboard](https://polydev.ai/dashboard)\n- **npm:** [npmjs.com/package/polydev-ai](https://www.npmjs.com/package/polydev-ai)\n- **Research:** [SWE-bench Paper](https://polydev.ai/articles/swe-bench-paper)\n\n### IDE Guides\n\n- [Claude Code Guide](https://polydev.ai/articles/claude-code-guide)\n- [Cursor Guide](https://polydev.ai/articles/cursor-guide)\n- [Cline Guide](https://polydev.ai/articles/cline-guide)\n- [Codex Guide](https://polydev.ai/articles/codex-guide)\n\n---\n\n## License\n\nMIT License - see [LICENSE](LICENSE) for details.\n\n---\n\n<p align=\"center\">\n  <b>Built by <a href=\"https://polydev.ai\">Polydev AI</a></b><br>\n  <i>Multi-model consultation for better code</i>\n</p>\n",
  "bytes": 5871,
  "sha": "a8d7704ff83ad8111d5a00ab17ccc366d12e423049c5606ce70a65df72d96e80",
  "repo_slug": "polydev-ai/perspectives-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_backspacevenkat_perspectives_2eaafe80/readme"
}