Vibe Coding

llama.cpp

benchmarks

GTX 6gig qwen2.5-coder-1.5b

MODEL=qwen2.5-coder-1.5b-instruct-q4_k_m.gguf
llama-server -m $MODEL \
  --host 0.0.0.0 \
  --port 9011 \
  -ngl 30 -c 16384 --jinja --flash-attn on

qwen2.5-coder-1.5b-instruct-q4_k_m.gguf Reading Generation 1,159 tokens 8.48s 136.62 tokens/s

qwen2.5-coder-7b

llama-server -m GLM-4.7-Flash-Q4_K_M.gguf  \
  --host 0.0.0.0 \
  --port 9011 \
  -ngl 8 -c 16384 --jinja --flash-attn on

qwen2.5-coder-7b-instruct-q4_k_m.gguf Reading Generation 1,968 tokens 330.68s 5.95 tokens/s

llama-server -m GLM-4.7-Flash-Q4_K_M.gguf  \
  --host 0.0.0.0 \
  --port 9011 \
  -ngl 13 -c 16384 --jinja --flash-attn on

Claude Code

MCP Server

https://docs.anthropic.com/en/docs/claude-code/mcp#add-mcp-servers-from-json-configuration

Add an MCP server from JSON

# Basic syntax
claude mcp add-json <name> '<json>'

# Example: Adding a stdio server with JSON configuration
claude mcp add-json weather-api '{"type":"stdio","command":"/path/to/weather-cli","args":["--api-key","abc123"],"env":{"CACHE_DIR":"/tmp"}}'

Verify the server was added

claude mcp get weather-api
weather-api:
  Scope: Local (private to you in this project)
  Type: stdio
  Command: /path/to/weather-cli
  Args: --api-key abc123
  Environment:
    CACHE_DIR=/tmp

To remove this server, run: claude mcp remove "weather-api" -s local

https://www.npmjs.com/package/@modelcontextprotocol/server-puppeteer

Environment Variable: Set PUPPETEER_LAUNCH_OPTIONS with a JSON-encoded string in the MCP configuration's env parameter:

{
  "mcpServers": {
    "mcp-puppeteer": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-puppeteer"],
      "env": {
        "PUPPETEER_LAUNCH_OPTIONS": "{ \"headless\": false, \"executablePath\": \"C:/Program Files/Google/Chrome/Application/chrome.exe\", \"args\": [] }",
        "ALLOW_DANGEROUS": "true"
      }
    }
  }
}

Other

https://github.com/1rgs/claude-code-proxy

git clone https://github.com/1rgs/claude-code-openai.git
cd claude-code-openai

curl -LsSf https://astral.sh/uv/install.sh | sh

uv run uvicorn server:app --host 0.0.0.0 --port 8082 --reload


OPENAI_API_KEY="your-openai-key"
GEMINI_API_KEY="your-google-key"
PREFERRED_PROVIDER="openai"
BIG_MODEL="gpt-4o" # Example specific model
SMALL_MODEL="gpt-4o-mini" # Example specific model

https://github.com/badlogic/lemmy/tree/main/apps/claude-bridge

npm install -g @mariozechner/claude-bridge

# Set API keys (optional - can specify per-command with --apiKey)
export OPENAI_API_KEY=sk-...
export GOOGLE_API_KEY=...

# Discovery workflow
claude-bridge                     # Show available providers
claude-bridge openai              # Show OpenAI models
claude-bridge openai gpt-4o       # Run Claude Code with GPT-4

# Advanced usage
claude-bridge openai gpt-4o --apiKey sk-...                           # Custom API key
claude-bridge openai llama3.2 --baseURL http://localhost:11434/v1     # Local Ollama
claude-bridge openai gpt-4o --baseURL https://openrouter.ai/api/v1 --apiKey sk-or-... # OpenRouter
claude-bridge openai gpt-4o --debug                          # Enable debug logs
claude-bridge --trace -p "Hello world"                       # Spy on Claude ↔ Anthropic communication

# All Claude Code arguments work
claude-bridge google gemini-2.5-pro-preview-05-06 --resume --continue
claude-bridge openai o4-mini -p "Hello world"