docs(spec): align AIProviderInterface with implementation (generate_changes/stream_changes)

Relates to #5801

The spec defined AIProviderInterface as a model factory with create_chat_model()
and create_embedding_model() methods. The implementation uses a higher-level
plan-execution interface with generate_changes() and stream_changes().

The implementation's approach is more appropriate — it directly handles plan
execution rather than exposing raw LangChain model creation. The ProviderRegistry
manages provider selection and configuration at a higher level.

Updated AIProviderInterface to match the actual implementation:
- provider_name -> name (property)
- capabilities -> model_id (property)
- create_chat_model() -> generate_changes() (plan execution)
- create_embedding_model() -> stream_changes() (streaming plan execution)

Also updated the description: auto-discovery of langchain-* packages is not
implemented; instead, providers are discovered based on configured API keys.

docs/specification.md

@@ -46427,31 +46427,52 @@ Actors can be extended through:
 
 1. **YAML-defined agents**: Single LLM actors with custom system prompts, temperature settings, tool bindings, and capability constraints.
 2. **YAML-defined graphs**: LangGraph topologies with multiple actors and tool nodes connected by edges, conditional routing, and parallel execution groups.
-3. **Provider extension**: New LLM providers can be added by implementing the `AIProviderInterface` protocol and registering with the `ProviderRegistry`. The registry uses auto-discovery to detect installed `langchain-*` packages.
+3. **Provider extension**: New LLM providers can be added by implementing the `AIProviderInterface` protocol and registering with the `ProviderRegistry`. The registry discovers configured providers based on available API keys and environment variables.
 
-<div class="highlight"><pre><code><span style="color: magenta; font-weight: 600;">from</span> typing <span style="color: magenta; font-weight: 600;">import</span> <span style="color: cyan;">Protocol</span>
+<div class="highlight"><pre><code><span style="color: magenta; font-weight: 600;">from</span> collections.abc <span style="color: magenta; font-weight: 600;">import</span> <span style="color: cyan;">Callable, Iterator</span>
+<span style="color: magenta; font-weight: 600;">from</span> typing <span style="color: magenta; font-weight: 600;">import</span> <span style="color: cyan;">Protocol</span>
 
 <span style="color: magenta; font-weight: 600;">class</span> <span style="color: cyan; font-weight: 600;">AIProviderInterface</span>(<span style="color: cyan;">Protocol</span>):
-    <span style="color: #888;">"""Protocol for LLM provider implementations."""</span>
+    <span style="color: #888;">"""Protocol for AI providers that generate code changes.
+
+    Implementations handle plan execution directly, generating changes
+    based on the plan, project context, and actor configuration.
+    Supported providers: openai, anthropic, google, gemini, azure,
+    openrouter, cohere, groq, together.
+    """</span>
 
     <span style="color: cyan;">@property</span>
-    <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">provider_name</span>(<span style="color: cyan;">self</span>) -> <span style="color: cyan;">str</span>: ...
+    <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">name</span>(<span style="color: cyan;">self</span>) -> <span style="color: cyan;">str</span>: ...
+    <span style="color: #888;">"""Provider name (e.g., 'openai', 'anthropic')."""</span>
 
     <span style="color: cyan;">@property</span>
-    <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">capabilities</span>(<span style="color: cyan;">self</span>) -> ProviderCapabilities: ...
+    <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">model_id</span>(<span style="color: cyan;">self</span>) -> <span style="color: cyan;">str</span>: ...
+    <span style="color: #888;">"""Model identifier (e.g., 'gpt-4o', 'claude-3-5-sonnet-20241022')."""</span>
 
-    <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">create_chat_model</span>(
+    <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">generate_changes</span>(
         <span style="color: cyan;">self</span>,
-        model: <span style="color: cyan;">str</span>,
-        temperature: <span style="color: cyan;">float</span> = <span style="color: yellow;">0.7</span>,
-        **kwargs,
-    ) -> BaseChatModel: ...
+        project: Project,
+        plan: Plan,
+        contexts: list[Context],
+        actor_context: ActorInvocationContext | None = None,
+        progress_callback: Callable[[int], None] | None = None,
+    ) -> ProviderResponse: ...
+    <span style="color: #888;">"""Generate code changes based on the plan and context."""</span>
 
-    <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">create_embedding_model</span>(
+    <span style="color: magenta; font-weight: 600;">def</span> <span style="color: cyan; font-weight: 600;">stream_changes</span>(
         <span style="color: cyan;">self</span>,
-        model: <span style="color: cyan;">str</span>,
-        **kwargs,
-    ) -> BaseEmbeddings: ...
+        project: Project,
+        plan: Plan,
+        contexts: list[Context],
+        actor_context: ActorInvocationContext | None = None,
+        progress_callback: Callable[[int], None] | None = None,
+    ) -> Iterator[dict[str, object]]: ...
+    <span style="color: #888;">"""Stream workflow events while generating changes.
+
+    Yields dicts keyed by workflow node name. Finishes with a
+    ``"__end__"`` event containing a ``ProviderResponse`` under
+    the ``response`` key.
+    """</span>
 </code></pre></div>
 
 #### Custom Resource Types