chore(pmoves): DRY make targets + Ollama URL

docs/PMOVES.AI-Edition-Hardened-Full.md

@@ -1024,7 +1024,7 @@ routing = ["ollama_local"]
 
 [models.qwen2_5_14b.providers.ollama_local]
 type = "openai"
-api_base = "http://pmoves-ollama:11434/v1"
+api_base = "http://ollama:11434/v1"
 model_name = "qwen2.5:14b"
 api_key_location = "none"
 
@@ -1033,7 +1033,7 @@ routing = ["ollama_local_embedding"]
 
 [embedding_models.qwen3_embedding_4b_local.providers.ollama_local_embedding]
 type = "openai"
-api_base = "http://pmoves-ollama:11434/v1"
+api_base = "http://ollama:11434/v1"
 model_name = "qwen3-embedding:4b"
 api_key_location = "none"
 ```

pmoves/Makefile

@@ -182,7 +182,7 @@ archon-submodule-extract: ## Extract Archon service to a submodule repo (set ARC
 
 .PHONY: up-archon-submodule
 up-archon-submodule: ## Build Archon from submodule (pmoves/integrations/archon)
-	@docker compose -p $(PROJECT) -f docker-compose.yml -f docker-compose.archon.submodule.yml up -d archon
+	@$(DC) up -d archon
 
 # -------- Consciousness Taxonomy Loaders ----------
 .PHONY: load-consciousness-neo4j harvest-consciousness
@@ -239,7 +239,7 @@ up: ensure-env-shared ## Start core data + workers and both Hi-RAG gateways
 	@echo "✔ Stack started (v2 on :$(HIRAG_CPU_PORT), v2-gpu on :$(HIRAG_GPU_PORT) when available)."
 
 up-gpu: ## Start with optional GPU accelerations where supported
-	@$(LOAD_ENV_SHARED) docker compose -f docker-compose.yml -f docker-compose.gpu.yml --profile gpu up -d
+	@$(DC) -f docker-compose.gpu.yml --profile gpu up -d
 	@echo "✔ Stack started with GPU profile."
 
 .PHONY: up-gpu-gateways

pmoves/docs/pmoves-model-management-starter/README.md

@@ -3,7 +3,7 @@
 This starter explains how to pick embedding/rerank models, switch providers at runtime, and bring up Agent Zero and Archon UIs for orchestration.
 
 ## Embeddings
-- Local (Ollama): set `USE_OLLAMA_EMBED=true`, `OLLAMA_URL=http://pmoves-ollama:11434`, and pick `OLLAMA_EMBED_MODEL=qwen3-embedding:4b` (Jetson/low VRAM: `qwen3-embedding:0.6b` or `embeddinggemma:300m`).
+- Local (Ollama): set `USE_OLLAMA_EMBED=true`, `OLLAMA_URL=http://ollama:11434`, and pick `OLLAMA_EMBED_MODEL=qwen3-embedding:4b` (Jetson/low VRAM: `qwen3-embedding:0.6b` or `embeddinggemma:300m`).
 - Remote (TensorZero): set `EMBEDDING_BACKEND=tensorzero`, `TENSORZERO_BASE_URL=http://<remote>:3000`, and choose `TENSORZERO_EMBED_MODEL` (default `tensorzero::embedding_model_name::qwen3_embedding_4b_local`).
 - Fallback: without providers, hi-rag uses `all-MiniLM-L6-v2` via sentence-transformers.
 

pmoves/docs/venice-tensorzero-integration/README.md

@@ -18,7 +18,7 @@ This guide shows how to use PMOVES with a local or remote TensorZero gateway and
   - `TENSORZERO_EMBED_MODEL=tensorzero::embedding_model_name::qwen3_embedding_4b_local`
 - Ollama backend:
   - `USE_OLLAMA_EMBED=true`
-  - `OLLAMA_URL=http://pmoves-ollama:11434`
+  - `OLLAMA_URL=http://ollama:11434`
   - `OLLAMA_EMBED_MODEL=qwen3-embedding:4b`
 - Fallback: if neither provider is reachable, hi-rag uses `SentenceTransformer` (`all-MiniLM-L6-v2`).
 

pmoves/env.shared.example

@@ -107,7 +107,7 @@ TENSORZERO_CLICKHOUSE_DB=tensorzero
 LANGEXTRACT_PROVIDER=rule
 
 # Ollama local models. `make up-tensorzero` also launches a bundled Ollama sidecar.
-OLLAMA_URL=http://pmoves-ollama:11434
+OLLAMA_URL=http://ollama:11434
 # Production default (RTX 5090 class): Qwen3-Embedding 4B. For edge/Jetson, prefer `qwen3-embedding:0.6b` or `embeddinggemma:300m`.
 OLLAMA_EMBED_MODEL=qwen3-embedding:4b