feat(v2): add Parse() function and extensible Registry for model string resolution

Introduces llm.Parse(spec) backed by an extensible Registry that resolves
model strings like "openai/gpt-4o", aliases like "fast", and named targets
like "m5/qwen3:30b" (via LLM_M5 env var DSNs) into ready-to-use *Model
objects. Extension points: RegisterProvider, RegisterAlias, RegisterResolver.
Adds Foreman constructor and sentinel errors ErrAliasLoop, ErrUnknownProvider,
ErrInvalidDSN.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

v2/constructors.go

@@ -119,6 +119,22 @@ func Ollama(opts ...ClientOption) *Client {
 	return NewClient(ollamaProvider.New("", cfg.baseURL))
 }
 
+// Foreman creates a client targeting a foreman daemon (a private, authenticated
+// Ollama endpoint with queuing and observability). The token is sent as a Bearer
+// token; pass "" for unauthenticated (network-trusted) deployments. Use
+// WithBaseURL to set the foreman host URL.
+//
+// Example:
+//
+//	model := llm.Foreman("my-token", llm.WithBaseURL("https://foreman.local")).Model("qwen3:30b")
+func Foreman(token string, opts ...ClientOption) *Client {
+	cfg := &clientConfig{}
+	for _, opt := range opts {
+		opt(cfg)
+	}
+	return NewClient(ollamaProvider.New(token, cfg.baseURL))
+}
+
 // OllamaCloud creates a client targeting Ollama Cloud (https://ollama.com).
 // The apiKey is required and is sent as `Authorization: Bearer <key>`. Use
 // WithBaseURL to point at a private Ollama deployment that requires auth.