feat: correx-managed model lifecycle slice 1 — config + manager factory + boot/shutdown

[[models]] config (ModelConfig/ModelsSettings) parsed by ConfigLoader;
InfrastructureModule.createModelManager + modelConfigToDescriptor; Main.kt
managed boot path spawns the default llama-server when [[models]] is present
(static [[providers]] path preserved when absent) and kills it on shutdown.

Plan: docs/plans/2026-05-31-model-lifecycle-management.md (slice 1 of 5).

docs/sample-config.toml

@@ -30,23 +30,48 @@ enabled = true
 [tools.file_edit]
 enabled = true
 
-# Provider configuration (array of tables)
-[[providers]]
-id = "local-llama"
-type = "llamacpp"
-model_id = "mistral-7b"
+# ─────────────────────────────────────────────────────────────
+# Managed model configuration (Slice 1: correx spawns + owns llama-server)
+#
+# When [[models]] is present, correx launches llama-server at boot and kills it on
+# shutdown.  Use this instead of (or alongside) [[providers]].
+#
+# [models] — global settings for the managed llama-server process
+# [[models]] — one entry per model file; correx will load the defaultModel at startup.
+# ─────────────────────────────────────────────────────────────
+[models]
+default_model = "mistral-7b"          # which [[models]] entry to load at boot
+llama_server_bin = "llama-server"     # path to the llama-server binary (default: llama-server)
+host = "127.0.0.1"                    # host for llama-server to bind / correx to connect
+port = 10000                          # port for llama-server
+
+[[models]]
+id = "mistral-7b"
 model_path = "~/models/mistral-7b-gguf/model.gguf"
-url = "http://127.0.0.1:10000"
+context_size = 8192
 capabilities = { General = 1.0, Coding = 0.7, Reasoning = 0.6, Summarization = 0.8, ToolCalling = 0.5 }
 
-# Example: second provider (if you have multiple)
+# Example: additional model (swap via TUI in a later slice)
+# [[models]]
+# id = "codellama-7b"
+# model_path = "~/models/codellama-7b-gguf/model.gguf"
+# context_size = 4096
+# capabilities = { General = 0.8, Coding = 1.0, Reasoning = 0.7, Summarization = 0.6, ToolCalling = 0.7 }
+
+# ─────────────────────────────────────────────────────────────
+# Legacy static provider configuration (array of tables)
+# Use [[providers]] when correx should connect to an already-running llama-server
+# (i.e. you launch the server yourself externally).
+# If [[models]] is configured, [[providers]] entries are registered as additional
+# static providers alongside the managed one.
+# ─────────────────────────────────────────────────────────────
 # [[providers]]
-# id = "alternative-llama"
+# id = "local-llama"
 # type = "llamacpp"
-# model_id = "neural-chat-7b"
-# model_path = "~/models/neural-chat-7b-gguf/model.gguf"
-# url = "http://127.0.0.1:10001"
-# capabilities = { General = 0.9, Coding = 0.8, Reasoning = 0.7, Summarization = 0.75, ToolCalling = 0.6 }
+# model_id = "mistral-7b"
+# model_path = "~/models/mistral-7b-gguf/model.gguf"
+# url = "http://127.0.0.1:10000"
+# capabilities = { General = 1.0, Coding = 0.7, Reasoning = 0.6, Summarization = 0.8, ToolCalling = 0.5 }
 
 # Router configuration (optional, defaults shown below)
 [router.embedder]