1 line
22 KiB
JSON
1 line
22 KiB
JSON
{"repo": "bdougie/vibe-check", "files": {"docs/continue-config.md": {"596df6f1c17835e983d34adcb2e44f0695e7a95d": {"t": 1785747191.6038692, "n": 6970, "h": "f0b75672cf221008", "c": "# Continue Configuration Generator\n\nThe Continue Configuration Generator automatically creates a `config.yaml` file for the [Continue](https://continue.dev) VS Code extension based on available models and user preferences.\n\n## Overview\n\nContinue is an AI-powered coding assistant that integrates with VS Code. The configuration generator:\n\n- Detects locally installed Ollama models\n- Adds templates for commercial models (OpenAI, Anthropic)\n- Configures context providers and commands\n- Sets up autocomplete preferences\n- Saves configuration to `~/.continue/config.yaml`\n\n## Usage\n\n### Running the Generator\n\n```bash\nuv run benchmark/continue_config_generator.py\n```\n\nThe generator will:\n\n1. **Check for existing configuration** - Prompts to overwrite if one exists\n2. **Detect Ollama models** - Automatically finds all locally installed models\n3. **Configure commercial models** - Optionally add API keys for OpenAI/Anthropic\n4. **Generate complete config** - Creates a ready-to-use configuration file\n5. **Save to ~/.continue** - Places the config in the correct location\n\n### Interactive Prompts\n\nDuring setup, you'll be asked:\n\n```\nDo you have an OpenAI API key to configure? (y/n): \nEnter your OpenAI API key: \n\nDo you have an Anthropic API key to configure? (y/n):\nEnter your Anthropic API key: \n```\n\nYou can skip these and add API keys later by editing the config file directly.\n\n## Configuration Structure\n\n### Generated config.yaml\n\n```yaml\nname: Continue Configuration\nversion: 1.0\nschema: 1.0.0\n\nmodels:\n # Ollama models (auto-detected)\n - name: Ollama codestral\n provider: ollama\n model: codestral\n roles: \n - chat\n - autocomplete\n - edit\n autocompleteOptions:\n debounceDelay: 250\n maxPromptTokens: 2048\n prefixPercentage: 0.5\n\n # Commercial models\n - name: GPT-4\n provider: openai\n model: gpt-4\n apiKey: YOUR_OPENAI_API_KEY\n roles: \n - chat\n - edit\n defaultCompletionOptions:\n temperature: 0.7\n maxTokens: 2048\n\n - name: Claude 3.5 Sonnet\n provider: anthropic\n model: claude-3-5-sonnet-20241022\n apiKey: YOUR_ANTHROPIC_API_KEY\n roles: \n - chat\n - edit\n defaultCompletionOptions:\n temperature: 0.7\n maxTokens: 4096\n\n# Autocomplete settings\ntabAutocompleteModel:\n title: Ollama codestral\n provider: ollama\n model: codestral\n\n# Context providers\ncontextProviders:\n - name: code\n - name: docs\n - name: diff\n - name: terminal\n - name: problems\n - name: folder\n - name: codebase\n\n# Slash commands\nslashCommands:\n - name: edit\n description: Edit selected code\n - name: comment\n description: Write comments for the selected code\n - name: share\n description: Export the current chat session to markdown\n - name: cmd\n description: Generate a shell command\n - name: commit\n description: Generate a git commit message\n\n# Custom commands\ncustomCommands:\n - name: test\n prompt: Write comprehensive unit tests for the selected code\n description: Generate unit tests\n - name: check\n prompt: Review the selected code for potential bugs, security issues, and improvements\n description: Perform code review\n - name: benchmark\n prompt: Analyze the performance of the selected code and suggest optimizations\n description: Performance analysis\n```\n\n## Model Detection\n\n### Ollama Models\n\nThe generator automatically detects Ollama models by running `ollama list`. Models are categorized by their capabilities:\n\n- **Code-focused models** (codestral, deepseek-coder): Enabled for chat, autocomplete, and edit\n- **General models** (llama2, mistral): Enabled for chat and edit\n- **Other models**: Enabled for chat only\n\nThe first code-focused model found is set as the default autocomplete model.\n\n### Commercial Models\n\nTemplates are added for:\n\n- **OpenAI**: GPT-4 configuration\n- **Anthropic**: Claude 3.5 Sonnet configuration\n\nAPI keys can be added during setup or manually edited later.\n\n## Manual Configuration\n\n### Adding API Keys\n\nEdit `~/.continue/config.yaml` and replace placeholder keys:\n\n```yaml\n# Replace these with your actual keys\napiKey: YOUR_OPENAI_API_KEY → apiKey: sk-your-actual-key\napiKey: YOUR_ANTHROPIC_API_KEY → apiKey: sk-ant-your-actual-key\n```\n\n### Adding New Models\n\nAdd entries to the `models` section:\n\n```yaml\nmodels:\n - name: Custom Model\n provider: provider_name\n model: model_name\n apiKey: your_api_key # if required\n roles: \n - chat\n - edit\n```\n\n### Changing Autocomplete Model\n\nModify the `tabAutocompleteModel` section:\n\n```yaml\ntabAutocompleteModel:\n title: Your Preferred Model\n provider: provider_name\n model: model_name\n```\n\n## File Locations\n\n- **Configuration**: `~/.continue/config.yaml`\n- **Backup** (if exists): `~/.continue/config.yaml.backup`\n- **Generator script**: `benchmark/continue_config_generator.py`\n\n## Troubleshooting\n\n### Ollama Not Detected\n\nIf Ollama models aren't detected:\n\n1. Ensure Ollama is installed: `ollama --version`\n2. Check Ollama service is running: `ollama list`\n3. Pull models if needed: `ollama pull codestral`\n\n### Configuration Not Loading\n\nAfter generating the configuration:\n\n1. Restart VS Code\n2. Reload the Continue extension\n3. Check Continue output panel for errors\n\n### Invalid API Keys\n\nIf API keys are invalid:\n\n1. Check for typos in the keys\n2. Ensure keys have proper permissions\n3. Verify billing/credits are available\n\n## Dependencies\n\nThe configuration generator requires:\n\n- Python 3.8+\n- PyYAML (`uv pip install pyyaml`)\n- Ollama (optional, for local models)\n\n## Integration with Benchmarks\n\nThe Continue configuration integrates with the vibe-check benchmark framework:\n\n1. Configure Continue with your preferred models\n2. Run benchmarks to test model performance\n3. Use benchmark results to optimize model selection\n4. Update Continue config based on benchmark findings\n\n## Related Documentation\n\n- [Continue Documentation](https://docs.continue.dev/reference)\n- [Ollama Setup Guide](ollama.md)\n- [Model Selection Guide](models.md)\n- [Benchmark Task Runner](../README.md#running-benchmarks)\n\n## Example Workflow\n\n1. **Install Ollama and pull models**:\n ```bash\n ollama pull codestral\n ollama pull llama2\n ```\n\n2. **Run the configuration generator**:\n ```bash\n uv run benchmark/continue_config_generator.py\n ```\n\n3. **Add API keys** (if using commercial models):\n - Enter during setup, or\n - Edit `~/.continue/config.yaml` manually\n\n4. **Restart VS Code** to load the new configuration\n\n5. **Test with Continue**:\n - Open a code file\n - Use `Cmd+L` (Mac) or `Ctrl+L` (Windows/Linux) to open Continue chat\n - Try autocomplete with Tab\n - Use slash commands like `/edit` or `/comment`\n\n## Security Notes\n\n- API keys are stored in plain text in the config file\n- Ensure `~/.continue/config.yaml` has appropriate file permissions\n- Never commit config files with real API keys to version control\n- Consider using environment variables for sensitive keys in production"}, "c98557c5dcd67b5da33106d26a58c1f225e6a4bc": {"t": 1785771223.9509592, "n": 6970, "h": "f0b75672cf221008", "c": "# Continue Configuration Generator\n\nThe Continue Configuration Generator automatically creates a `config.yaml` file for the [Continue](https://continue.dev) VS Code extension based on available models and user preferences.\n\n## Overview\n\nContinue is an AI-powered coding assistant that integrates with VS Code. The configuration generator:\n\n- Detects locally installed Ollama models\n- Adds templates for commercial models (OpenAI, Anthropic)\n- Configures context providers and commands\n- Sets up autocomplete preferences\n- Saves configuration to `~/.continue/config.yaml`\n\n## Usage\n\n### Running the Generator\n\n```bash\nuv run benchmark/continue_config_generator.py\n```\n\nThe generator will:\n\n1. **Check for existing configuration** - Prompts to overwrite if one exists\n2. **Detect Ollama models** - Automatically finds all locally installed models\n3. **Configure commercial models** - Optionally add API keys for OpenAI/Anthropic\n4. **Generate complete config** - Creates a ready-to-use configuration file\n5. **Save to ~/.continue** - Places the config in the correct location\n\n### Interactive Prompts\n\nDuring setup, you'll be asked:\n\n```\nDo you have an OpenAI API key to configure? (y/n): \nEnter your OpenAI API key: \n\nDo you have an Anthropic API key to configure? (y/n):\nEnter your Anthropic API key: \n```\n\nYou can skip these and add API keys later by editing the config file directly.\n\n## Configuration Structure\n\n### Generated config.yaml\n\n```yaml\nname: Continue Configuration\nversion: 1.0\nschema: 1.0.0\n\nmodels:\n # Ollama models (auto-detected)\n - name: Ollama codestral\n provider: ollama\n model: codestral\n roles: \n - chat\n - autocomplete\n - edit\n autocompleteOptions:\n debounceDelay: 250\n maxPromptTokens: 2048\n prefixPercentage: 0.5\n\n # Commercial models\n - name: GPT-4\n provider: openai\n model: gpt-4\n apiKey: YOUR_OPENAI_API_KEY\n roles: \n - chat\n - edit\n defaultCompletionOptions:\n temperature: 0.7\n maxTokens: 2048\n\n - name: Claude 3.5 Sonnet\n provider: anthropic\n model: claude-3-5-sonnet-20241022\n apiKey: YOUR_ANTHROPIC_API_KEY\n roles: \n - chat\n - edit\n defaultCompletionOptions:\n temperature: 0.7\n maxTokens: 4096\n\n# Autocomplete settings\ntabAutocompleteModel:\n title: Ollama codestral\n provider: ollama\n model: codestral\n\n# Context providers\ncontextProviders:\n - name: code\n - name: docs\n - name: diff\n - name: terminal\n - name: problems\n - name: folder\n - name: codebase\n\n# Slash commands\nslashCommands:\n - name: edit\n description: Edit selected code\n - name: comment\n description: Write comments for the selected code\n - name: share\n description: Export the current chat session to markdown\n - name: cmd\n description: Generate a shell command\n - name: commit\n description: Generate a git commit message\n\n# Custom commands\ncustomCommands:\n - name: test\n prompt: Write comprehensive unit tests for the selected code\n description: Generate unit tests\n - name: check\n prompt: Review the selected code for potential bugs, security issues, and improvements\n description: Perform code review\n - name: benchmark\n prompt: Analyze the performance of the selected code and suggest optimizations\n description: Performance analysis\n```\n\n## Model Detection\n\n### Ollama Models\n\nThe generator automatically detects Ollama models by running `ollama list`. Models are categorized by their capabilities:\n\n- **Code-focused models** (codestral, deepseek-coder): Enabled for chat, autocomplete, and edit\n- **General models** (llama2, mistral): Enabled for chat and edit\n- **Other models**: Enabled for chat only\n\nThe first code-focused model found is set as the default autocomplete model.\n\n### Commercial Models\n\nTemplates are added for:\n\n- **OpenAI**: GPT-4 configuration\n- **Anthropic**: Claude 3.5 Sonnet configuration\n\nAPI keys can be added during setup or manually edited later.\n\n## Manual Configuration\n\n### Adding API Keys\n\nEdit `~/.continue/config.yaml` and replace placeholder keys:\n\n```yaml\n# Replace these with your actual keys\napiKey: YOUR_OPENAI_API_KEY → apiKey: sk-your-actual-key\napiKey: YOUR_ANTHROPIC_API_KEY → apiKey: sk-ant-your-actual-key\n```\n\n### Adding New Models\n\nAdd entries to the `models` section:\n\n```yaml\nmodels:\n - name: Custom Model\n provider: provider_name\n model: model_name\n apiKey: your_api_key # if required\n roles: \n - chat\n - edit\n```\n\n### Changing Autocomplete Model\n\nModify the `tabAutocompleteModel` section:\n\n```yaml\ntabAutocompleteModel:\n title: Your Preferred Model\n provider: provider_name\n model: model_name\n```\n\n## File Locations\n\n- **Configuration**: `~/.continue/config.yaml`\n- **Backup** (if exists): `~/.continue/config.yaml.backup`\n- **Generator script**: `benchmark/continue_config_generator.py`\n\n## Troubleshooting\n\n### Ollama Not Detected\n\nIf Ollama models aren't detected:\n\n1. Ensure Ollama is installed: `ollama --version`\n2. Check Ollama service is running: `ollama list`\n3. Pull models if needed: `ollama pull codestral`\n\n### Configuration Not Loading\n\nAfter generating the configuration:\n\n1. Restart VS Code\n2. Reload the Continue extension\n3. Check Continue output panel for errors\n\n### Invalid API Keys\n\nIf API keys are invalid:\n\n1. Check for typos in the keys\n2. Ensure keys have proper permissions\n3. Verify billing/credits are available\n\n## Dependencies\n\nThe configuration generator requires:\n\n- Python 3.8+\n- PyYAML (`uv pip install pyyaml`)\n- Ollama (optional, for local models)\n\n## Integration with Benchmarks\n\nThe Continue configuration integrates with the vibe-check benchmark framework:\n\n1. Configure Continue with your preferred models\n2. Run benchmarks to test model performance\n3. Use benchmark results to optimize model selection\n4. Update Continue config based on benchmark findings\n\n## Related Documentation\n\n- [Continue Documentation](https://docs.continue.dev/reference)\n- [Ollama Setup Guide](ollama.md)\n- [Model Selection Guide](models.md)\n- [Benchmark Task Runner](../README.md#running-benchmarks)\n\n## Example Workflow\n\n1. **Install Ollama and pull models**:\n ```bash\n ollama pull codestral\n ollama pull llama2\n ```\n\n2. **Run the configuration generator**:\n ```bash\n uv run benchmark/continue_config_generator.py\n ```\n\n3. **Add API keys** (if using commercial models):\n - Enter during setup, or\n - Edit `~/.continue/config.yaml` manually\n\n4. **Restart VS Code** to load the new configuration\n\n5. **Test with Continue**:\n - Open a code file\n - Use `Cmd+L` (Mac) or `Ctrl+L` (Windows/Linux) to open Continue chat\n - Try autocomplete with Tab\n - Use slash commands like `/edit` or `/comment`\n\n## Security Notes\n\n- API keys are stored in plain text in the config file\n- Ensure `~/.continue/config.yaml` has appropriate file permissions\n- Never commit config files with real API keys to version control\n- Consider using environment variables for sensitive keys in production"}, "6364106a726e77797ecf83aec43a5aeb6a3c765f": {"t": 1785771227.7858522, "n": 6967, "h": "2dd085c004578318", "c": "# Continue Configuration Generator\n\nThe Continue Configuration Generator automatically creates a `config.yaml` file for the [Continue](https://continue.dev) VS Code extension based on available models and user preferences.\n\n## Overview\n\nContinue is an AI-powered coding assistant that integrates with VS Code. The configuration generator:\n\n- Detects locally installed Ollama models\n- Adds templates for commercial models (OpenAI, Anthropic)\n- Configures context providers and commands\n- Sets up autocomplete preferences\n- Saves configuration to `~/.continue/config.yaml`\n\n## Usage\n\n### Running the Generator\n\n```bash\npython benchmark/continue_config_generator.py\n```\n\nThe generator will:\n\n1. **Check for existing configuration** - Prompts to overwrite if one exists\n2. **Detect Ollama models** - Automatically finds all locally installed models\n3. **Configure commercial models** - Optionally add API keys for OpenAI/Anthropic\n4. **Generate complete config** - Creates a ready-to-use configuration file\n5. **Save to ~/.continue** - Places the config in the correct location\n\n### Interactive Prompts\n\nDuring setup, you'll be asked:\n\n```\nDo you have an OpenAI API key to configure? (y/n): \nEnter your OpenAI API key: \n\nDo you have an Anthropic API key to configure? (y/n):\nEnter your Anthropic API key: \n```\n\nYou can skip these and add API keys later by editing the config file directly.\n\n## Configuration Structure\n\n### Generated config.yaml\n\n```yaml\nname: Continue Configuration\nversion: 1.0\nschema: 1.0.0\n\nmodels:\n # Ollama models (auto-detected)\n - name: Ollama codestral\n provider: ollama\n model: codestral\n roles: \n - chat\n - autocomplete\n - edit\n autocompleteOptions:\n debounceDelay: 250\n maxPromptTokens: 2048\n prefixPercentage: 0.5\n\n # Commercial models\n - name: GPT-4\n provider: openai\n model: gpt-4\n apiKey: YOUR_OPENAI_API_KEY\n roles: \n - chat\n - edit\n defaultCompletionOptions:\n temperature: 0.7\n maxTokens: 2048\n\n - name: Claude 3.5 Sonnet\n provider: anthropic\n model: claude-3-5-sonnet-20241022\n apiKey: YOUR_ANTHROPIC_API_KEY\n roles: \n - chat\n - edit\n defaultCompletionOptions:\n temperature: 0.7\n maxTokens: 4096\n\n# Autocomplete settings\ntabAutocompleteModel:\n title: Ollama codestral\n provider: ollama\n model: codestral\n\n# Context providers\ncontextProviders:\n - name: code\n - name: docs\n - name: diff\n - name: terminal\n - name: problems\n - name: folder\n - name: codebase\n\n# Slash commands\nslashCommands:\n - name: edit\n description: Edit selected code\n - name: comment\n description: Write comments for the selected code\n - name: share\n description: Export the current chat session to markdown\n - name: cmd\n description: Generate a shell command\n - name: commit\n description: Generate a git commit message\n\n# Custom commands\ncustomCommands:\n - name: test\n prompt: Write comprehensive unit tests for the selected code\n description: Generate unit tests\n - name: check\n prompt: Review the selected code for potential bugs, security issues, and improvements\n description: Perform code review\n - name: benchmark\n prompt: Analyze the performance of the selected code and suggest optimizations\n description: Performance analysis\n```\n\n## Model Detection\n\n### Ollama Models\n\nThe generator automatically detects Ollama models by running `ollama list`. Models are categorized by their capabilities:\n\n- **Code-focused models** (codestral, deepseek-coder): Enabled for chat, autocomplete, and edit\n- **General models** (llama2, mistral): Enabled for chat and edit\n- **Other models**: Enabled for chat only\n\nThe first code-focused model found is set as the default autocomplete model.\n\n### Commercial Models\n\nTemplates are added for:\n\n- **OpenAI**: GPT-4 configuration\n- **Anthropic**: Claude 3.5 Sonnet configuration\n\nAPI keys can be added during setup or manually edited later.\n\n## Manual Configuration\n\n### Adding API Keys\n\nEdit `~/.continue/config.yaml` and replace placeholder keys:\n\n```yaml\n# Replace these with your actual keys\napiKey: YOUR_OPENAI_API_KEY → apiKey: sk-your-actual-key\napiKey: YOUR_ANTHROPIC_API_KEY → apiKey: sk-ant-your-actual-key\n```\n\n### Adding New Models\n\nAdd entries to the `models` section:\n\n```yaml\nmodels:\n - name: Custom Model\n provider: provider_name\n model: model_name\n apiKey: your_api_key # if required\n roles: \n - chat\n - edit\n```\n\n### Changing Autocomplete Model\n\nModify the `tabAutocompleteModel` section:\n\n```yaml\ntabAutocompleteModel:\n title: Your Preferred Model\n provider: provider_name\n model: model_name\n```\n\n## File Locations\n\n- **Configuration**: `~/.continue/config.yaml`\n- **Backup** (if exists): `~/.continue/config.yaml.backup`\n- **Generator script**: `benchmark/continue_config_generator.py`\n\n## Troubleshooting\n\n### Ollama Not Detected\n\nIf Ollama models aren't detected:\n\n1. Ensure Ollama is installed: `ollama --version`\n2. Check Ollama service is running: `ollama list`\n3. Pull models if needed: `ollama pull codestral`\n\n### Configuration Not Loading\n\nAfter generating the configuration:\n\n1. Restart VS Code\n2. Reload the Continue extension\n3. Check Continue output panel for errors\n\n### Invalid API Keys\n\nIf API keys are invalid:\n\n1. Check for typos in the keys\n2. Ensure keys have proper permissions\n3. Verify billing/credits are available\n\n## Dependencies\n\nThe configuration generator requires:\n\n- Python 3.8+\n- PyYAML (`pip install pyyaml`)\n- Ollama (optional, for local models)\n\n## Integration with Benchmarks\n\nThe Continue configuration integrates with the vibe-check benchmark framework:\n\n1. Configure Continue with your preferred models\n2. Run benchmarks to test model performance\n3. Use benchmark results to optimize model selection\n4. Update Continue config based on benchmark findings\n\n## Related Documentation\n\n- [Continue Documentation](https://docs.continue.dev/reference)\n- [Ollama Setup Guide](ollama.md)\n- [Model Selection Guide](models.md)\n- [Benchmark Task Runner](../README.md#running-benchmarks)\n\n## Example Workflow\n\n1. **Install Ollama and pull models**:\n ```bash\n ollama pull codestral\n ollama pull llama2\n ```\n\n2. **Run the configuration generator**:\n ```bash\n python benchmark/continue_config_generator.py\n ```\n\n3. **Add API keys** (if using commercial models):\n - Enter during setup, or\n - Edit `~/.continue/config.yaml` manually\n\n4. **Restart VS Code** to load the new configuration\n\n5. **Test with Continue**:\n - Open a code file\n - Use `Cmd+L` (Mac) or `Ctrl+L` (Windows/Linux) to open Continue chat\n - Try autocomplete with Tab\n - Use slash commands like `/edit` or `/comment`\n\n## Security Notes\n\n- API keys are stored in plain text in the config file\n- Ensure `~/.continue/config.yaml` has appropriate file permissions\n- Never commit config files with real API keys to version control\n- Consider using environment variables for sensitive keys in production"}}}} |