CRUSH.md

Catwalk - AI Provider Database

Build/Test Commands

  • go run . - Build and run the main HTTP server on :8080
  • go run ./cmd/{provider-name} - Build and run a CLI to update the {provider-name}.json file
  • go test ./... - Run all tests

Code Style Guidelines

  • Package comments: Start with "Package name provides/represents..."
  • Imports: Standard library first, then third-party, then local packages
  • Error handling: Use fmt.Errorf("message: %w", err) for wrapping
  • Struct tags: Use json tags with omitempty for optional fields
  • Constants: Group related constants with descriptive comments
  • Types: Use custom types for IDs (e.g., InferenceProvider, Type)
  • Naming: Use camelCase for unexported, PascalCase for exported
  • Comments: Use //nolint:directive for linter exceptions
  • HTTP: Always set timeouts, use context, defer close response bodies
  • JSON: Use json.MarshalIndent for pretty output, validate unmarshaling
  • File permissions: Use 0o600 for sensitive config files
  • Always format code with gofumpt

Model Names

The internal/names package provides human-readable display names for model IDs. Only use this package when the provider API does not already provide a good display name.

Check the API response first:

  • If the API provides a name or display_name field, use that directly
  • Only use names.GetDisplayName(modelID) when you only have a model_id field

Examples:

  • Use API name: OpenRouter API has a name field → use model.Name
  • Use names package: AIHubMix API only has model_id → use names.GetDisplayName(model.ModelID)
import "github.com/charmbracelet/catwalk/internal/names"

// Only when API doesn't provide a name:
model := catwalk.Model{
    ID:   modelID,
    Name: names.GetDisplayName(modelID),
    // ... other fields
}

The names package uses:

  1. Static mappings for known models (most common models)
  2. Case-insensitive matching
  3. Provider prefix stripping (e.g., "anthropic/claude-sonnet-4" -> matches "claude-sonnet-4")
  4. Levenshtein distance fuzzy matching for unknown models
  5. Smart formatting for completely unknown models (converts "3-5" to "3.5", etc.)

To add new model mappings, edit internal/names/model.go and add entries to the modelNames map.

Adding more provider commands

  • Create the ./cmd/{provider-name}/main.go file
  • Try to use the provider API to figure out the available models. If there's no endpoint for listing the models, look for some sort of structured text format (usually in the docs). If none of that exist, refuse to create the command, and add it to the MANUAL_UPDATES.md file.
  • Add it to .github/workflows/update.yml

Updating providers manually

Zai

For zai, we'll need to grab the model list and capabilities from https://docs.z.ai/guides/overview/overview.

That page does not contain the exact context_window and default_max_tokens though. We can grab the exact value from ./internal/providers/configs/openrouter.json.