# PromptTuner MCP Configuration Guide
PromptTuner MCP uses minimal configuration. Set the provider and API key, and you're ready to go.
## Required Configuration
| Variable | Default | Description |
| ------------------- | -------- | ----------------------------------------------------- |
| `LLM_PROVIDER` | `openai` | `openai`, `anthropic`, or `google`. |
| `OPENAI_API_KEY` | - | Required for all tools when `LLM_PROVIDER=openai`. |
| `ANTHROPIC_API_KEY` | - | Required for all tools when `LLM_PROVIDER=anthropic`. |
| `GOOGLE_API_KEY` | - | Required for all tools when `LLM_PROVIDER=google`. |
All tools are LLM-backed and require an API key for the selected provider.
## Optional Configuration
| Variable | Default | Description |
| ---------------- | ------- | ------------------------------------------------ |
| `LLM_MODEL` | - | Override the default model. |
| `LLM_TIMEOUT_MS` | `15000` | Override per-request LLM timeout (milliseconds). |
| `DEBUG` | `false` | Enable debug logging. |
## Default Models
| Provider | Default Model |
| ----------- | ---------------------------- |
| `openai` | `gpt-4o` |
| `anthropic` | `claude-3-5-sonnet-20241022` |
| `google` | `gemini-2.0-flash-exp` |
## CLI Flags
| Flag | Description |
| --------------------------- | ----------------------------------- |
| `-h, --help` | Show help text. |
| `-v, --version` | Print version. |
| `--debug / --no-debug` | Enable/disable debug logging. |
| `--llm-provider <provider>` | `openai`, `anthropic`, or `google`. |
| `--llm-model <name>` | Override the default model. |
CLI flags override environment variables.
## Example Configurations
### Minimal (npx)
```json
{
"mcpServers": {
"prompttuner": {
"command": "npx",
"args": ["-y", "@j0hanz/prompt-tuner-mcp-server@latest"],
"env": {
"LLM_PROVIDER": "openai",
"OPENAI_API_KEY": "${input:openai-api-key}"
}
}
}
}
```
### With Custom Model
```json
{
"mcpServers": {
"prompttuner": {
"command": "npx",
"args": ["-y", "@j0hanz/prompt-tuner-mcp-server@latest"],
"env": {
"LLM_PROVIDER": "openai",
"OPENAI_API_KEY": "${input:openai-api-key}",
"LLM_MODEL": "gpt-4o-mini"
}
}
}
}
```
### Google Gemini
```json
{
"mcpServers": {
"prompttuner": {
"command": "npx",
"args": ["-y", "@j0hanz/prompt-tuner-mcp-server@latest"],
"env": {
"LLM_PROVIDER": "google",
"GOOGLE_API_KEY": "${input:google-api-key}"
}
}
}
}
```
## Default Settings
The following are optimized defaults:
| Setting | Value | Purpose |
| ------------------- | ------- | ------------------------------------------- |
| Max prompt length | 10,000 | Input character limit |
| Retry attempts | 3 | Retries on transient failures |
| Retry base delay | 1s | Exponential backoff base |
| Retry max delay | 10s | Backoff cap |
| Retry total timeout | 180s | Total retry window |
| Max output tokens | 10,000 | Output token cap (scales with input length) |
| Google safety | Enabled | Content filtering |
## Troubleshooting
### Prompt rejected
Shorten the prompt or remove excessive whitespace (limit: 10,000 chars).
### Timeout errors
Try a faster model like `gpt-4o-mini` or `gemini-2.0-flash-exp`.
### Rate limit errors
Reduce request frequency. The server automatically retries with backoff.
### Debug mode
Set `DEBUG=true` to see detailed logs including LLM request telemetry.
## Best Practices
1. Only configure the API key for your chosen provider.
2. Use input variables for secrets: `"OPENAI_API_KEY": "${input:openai-api-key}"`.
3. Start with defaults—they work well for most use cases.
4. Use `gpt-4o-mini` or `gemini-2.0-flash-exp` for faster, cheaper responses.
