Replace CLI codegen references with @constructive-io/graphql-codegen package + skill links

Author: pyramation

skills/constructive-ai/SKILL.md

@@ -23,7 +23,7 @@ Use this skill when:
 
 ```
 1. Provision  →  SDK creates vector(N) column + HNSW index on your table
-2. Codegen    →  cnc codegen --orm generates typed ORM with vector query support
+2. Codegen    →  @constructive-io/graphql-codegen generates typed ORM (see constructive-graphql skill)
 3. Embed      →  Application code generates embeddings (Ollama, OpenAI, etc.)
 4. Store      →  ORM or SDK inserts embeddings into the vector column
 5. Query      →  ORM queries with vectorEmbedding filter + distance ordering

skills/constructive-ai/references/rag-pipeline.md

@@ -15,7 +15,7 @@ Query -> Embedding -> ORM Vector Search -> Context Retrieval -> LLM Response
 ### Components
 
 1. **Provision**: SDK creates vector columns + HNSW indexes on your tables
-2. **Codegen**: `cnc codegen --orm` generates typed ORM with vector query support
+2. **Codegen**: `@constructive-io/graphql-codegen` generates typed ORM with vector query support (see `constructive-graphql` skill — [codegen.md](../../constructive-graphql/references/codegen.md))
 3. **Ingestion**: Embed documents with Ollama, store via ORM
 4. **Retrieval**: ORM `vectorEmbedding` filter + distance ordering
 5. **Generation**: Feed retrieved context to LLM for response
@@ -70,11 +70,7 @@ await db.index.create({
 
 ## Step 2: Generate ORM
 
-```bash
-cnc codegen --orm --react-query
-```
-
-This generates typed ORM methods including vector search support. See `constructive-graphql` skill ([codegen.md](../../constructive-graphql/references/codegen.md)) for full setup.
+Use `@constructive-io/graphql-codegen` to generate a typed ORM client with vector search support (ORM and React Query hooks are both available). See the `constructive-graphql` skill ([codegen.md](../../constructive-graphql/references/codegen.md)) for full setup and options.
 
 ---
 

skills/constructive-graphql/references/search-rag.md

@@ -6,7 +6,7 @@ Use the codegen'd ORM to build Retrieval-Augmented Generation (RAG) pipelines. T
 
 1. **Vector column** provisioned on your table (via `@constructive-io/sdk` — see `constructive-ai` skill)
 2. **HNSW index** created on the embedding column
-3. **Codegen'd SDK** generated with `cnc codegen --orm`
+3. **Codegen'd SDK** generated via `@constructive-io/graphql-codegen` (see `constructive-graphql` skill — [codegen.md](./codegen.md))
 4. **Ollama** (or another provider) running for embeddings and chat
 
 ## Vector Search via ORM