Skip to main content

Boring: An AI-powered development support tool for autonomous reasoning, planning, and verification. Optimized for uv and agentic workflows.

Project description

Boring for Gemini

PyPI version License

Autonomous AI Agent Loop with VibeCoder Experience


✨ The Vibe Coder Experience

No Code Needed. Just describe the vibe.

Boring-Gemini features a Universal Natural Language Router. You don't need to remember 60+ complex tools. Just say what you want:

"Search for authentication logic"
"Review my code for security issues"
"幫我寫測試" (Help me write tests)


🚀 Quick Start (One-Click)

🪄 Zero-Config Wizard

Boring features a smart setup wizard that automatically bridges your IDE (Cursor, VS Code, Claude Desktop) with the Gemini agent.

pip install "boring-aicoding[all]"
boring wizard

💎 Smithery Integration

One-click install for Cursor/Claude Desktop.

npx -y @smithery/cli@latest install boring/boring --client cursor

🔮 The Anti-Rage Revolution (V15.0.0)

V15.0.0 focuses on UX Hardening and Resilience, transforming Boring into a robust partner that doesn't just work, but communicates its progress:

  • 🧠 Cognitive Architecture (System 2): Uses sequentialthinking to plan, reason, and revise before execution.
  • 🐉 One Dragon Flow: Autonomous traversal of the entire SDLC (Design → Build → Polish → Verify).
  • 🔮 Predictive Intelligence: boring predict and boring bisect catch errors before they happen.
  • 🔌 Offline-First Mode: 100% network-independent operation with local LLMs via llama-cpp-python.
  • 🛡️ Sovereign Autonomy: Managed tool governance (boring policy) and knowledge continuity (boring bio).
  • 🩺 Self-Healing: boring doctor --fix automatically repairs your environment.

⚡ Core Capabilities

Feature Description
🧠 Autonomous Loop Thinking → Coding → Testing → Fixing with System 2 reasoning.
🕵️ Context Optimization 80% token savings via smart chunking and semantic specificity ranking.
🛡️ Shadow Mode Safe execution sandbox catches destructive operations before they happen.
🧩 Natural Language Universal Router: "幫我寫測試" or "Fix auth bugs" triggered instantly.
🧪 Vibe Score Gamified health check: Lint, Security, and Docs tracked over time.

📦 Usage Modes

As MCP Server (Recommended for IDEs)

{
  "mcpServers": {
    "boring": { "command": "boring-mcp" }
  }
}

Managed CLI Utilities

  • boring flow: Start the One Dragon autonomous flow.
  • boring wizard: Bridge Cursor, Claude, VSCode, and more in seconds.
  • boring-monitor: Fullscreen TUI dashboard for live agent monitoring.
  • boring-dashboard: Modern web UI for project health and history.

About boring-route: This command only shows which MCP tool would be selected for your query. To actually execute the tool, use it in your MCP client (Cursor/Claude Desktop).


🔧 Key Features

  • 67+ MCP Tools: RAG, Testing, Review, Security, Git, and more
  • Predictive Error Detection: pro-active anti-pattern analysis
  • Local LLM Support: Llama 3, Qwen 2.5, Phi-3 integration
  • Multi-Language: Python, JavaScript, TypeScript, Go, C++ support
  • Quality Gates: CI/CD integration with multi-tier verification
  • Memory System: Persistent learning and pattern recognition

📚 Documentation


📄 License

Apache 2.0 - See LICENSE

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

boring_aicoding-15.0.0.tar.gz (698.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

boring_aicoding-15.0.0-py3-none-any.whl (821.7 kB view details)

Uploaded Python 3

File details

Details for the file boring_aicoding-15.0.0.tar.gz.

File metadata

  • Download URL: boring_aicoding-15.0.0.tar.gz
  • Upload date:
  • Size: 698.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.14

File hashes

Hashes for boring_aicoding-15.0.0.tar.gz
Algorithm Hash digest
SHA256 4e48479dcf24fcba19ba7aefa71ef0a844cc77a26412aa01b09a53c416c8530a
MD5 59ba9ac474eec4d2887317b6fa220342
BLAKE2b-256 0546270d1650c1059af856a63d486a214e0c31c8542fadafc9aaa9a4c76751ec

See more details on using hashes here.

File details

Details for the file boring_aicoding-15.0.0-py3-none-any.whl.

File metadata

File hashes

Hashes for boring_aicoding-15.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 6c9883709d85b5215b3b4352c0c3f0c2a62117cad4e8a05503c10114041de1d8
MD5 85e4c96b7b5eb688cbc3430a5fc3863d
BLAKE2b-256 b4b6a20b6739733e7775d78bed56f7d5b3c619b0aa72b35d4d5a0e23fb976a11

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page