feat: Add local model runner support (ollama, vllm, llama.cpp) (#5)

* feat: Add local model runner support (ollama, vllm, llama.cpp)

Adds infrastructure for spawning and managing local LLM runners:

- **fetch.rs**: Generic file fetching for local paths and remote URLs
  with caching support and huggingface:// URL translation

- **ollama.rs**: Modelfile parsing and generation with parameter
  merging for runtime configuration

- **vllm.rs**: vLLM CLI argument generation for OpenAI-compatible
  server deployment

- **llamacpp.rs**: llama.cpp server CLI argument generation with
  parameter aliasing

- **runner.rs**: RunnerManager for spawning/stopping local model
  processes with graceful shutdown

- **ModelConfig**: Unified model configuration with runner field
  (external, ollama, vllm, llama-cpp) replacing the old
  ModelDefinition enum (backward compatible)

- **context**: Architecture nodes can specify deployment context
  override for remote deployment of local runners

Includes test fixture: tinyllama.Modelfile for Modelfile testing.

* Add runner management endpoints and remote deployment tests

- Add /v1/runners endpoints for spawn, list, and stop operations
- Extend AppState to include optional SharedRunnerManager
- Add with_runner_manager() builder for worker mode
- Create integration tests simulating control plane to worker dispatch
- Tests verify multi-instance coordination for remote deployment

This enables the control plane to dispatch spawn commands to workers
representing remote clusters, as discussed in the context override design.

* Add Docker runner support and improve vLLM integration

Docker support:
- Add docker.rs module with DockerConfig, RegistryConfig structs
- Support image-based and Dockerfile-based container deployment
- Handle custom registries with authentication
- Map model parameters to container environment variables
- Support GPU passthrough, volumes, network modes, IPC settings
- Add extra_args for vLLM-specific flags (--swap-space, --tool-call-parser)

vLLM improvements:
- Add is_vllm_installed() and is_cuda_available() checks
- Add HuggingFace token detection from HF_TOKEN env var
- Detect gated models requiring authentication
- Generate proper environment variables for vLLM process
- Add warnings for missing CUDA or HF token

Example Docker config matching user's DGX setup:
{
  "runner": "docker",
  "source": "RESMP-DEV/Qwen3-Next-80B-A3B-Instruct-NVFP4",
  "docker": {
    "image": "dgx-vllm:cutlass-nvfp4",
    "network": "host",
    "gpus": "all",
    "ipc": "host",
    "volumes": ["${HOME}/.cache/huggingface:/root/.cache/huggingface"],
    "extra_args": "--swap-space 32 --tool-call-parser hermes"
  },
  "parameters": {
    "tensor_parallel_size": 1,
    "max_model_len": 131072
  }
}

* Refactor extra_args to structured map for cleaner composability

- Change extra_args from string to HashMap<String, Value>
- Add extra_args_to_string() to convert map to CLI format at runtime
- Support bool flags (true=include, false=omit), numbers, strings, arrays
- Arrays repeat the flag for each value (useful for --stop tokens)

Example:
  "extra_args": {
    "swap-space": 32,
    "tool-call-parser": "hermes",
    "enable-auto-tool-choice": true
  }

Becomes: "--swap-space 32 --tool-call-parser hermes --enable-auto-tool-choice"

This cleans up the rough edges of CLI tools and makes composition files
more readable and programmatically manipulable.

Author: tbraun96

Cargo.lock

@@ -50,6 +50,9 @@ serde_yaml = "0.9"
 # Home directory detection
 dirs = "5"
 
+# Hashing for cache keys
+sha2 = "0.10"
+
 # System hostname
 hostname = "0.4"
 

Cargo.toml

@@ -35,6 +35,11 @@ pub struct ArchitectureNode {
     /// WebSocket URL for "ws" adapter
     pub url: Option<String>,
 
+    /// Deployment context override (default: localhost)
+    /// Use to deploy local runners to remote nodes
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub context: Option<String>,
+
     #[serde(rename = "extra-options", default)]
     pub extra_options: HashMap<String, serde_json::Value>,
 }
@@ -163,8 +168,23 @@ mod tests {
             use_case: None,
             condition: None,
             url: None,
+            context: None,
             extra_options: HashMap::new(),
         };
         assert_eq!(node.effective_bind_addr(), "0.0.0.0");
     }
+
+    #[test]
+    fn test_parse_node_with_context() {
+        let json = r#"{
+            "name": "remote-model",
+            "layer": 1,
+            "model": "llama-vllm",
+            "adapter": "openai-api",
+            "context": "gpu-cluster"
+        }"#;
+
+        let node: ArchitectureNode = serde_json::from_str(json).unwrap();
+        assert_eq!(node.context, Some("gpu-cluster".to_string()));
+    }
 }