diff --git a/.opencode/skills/rag-workflows/SKILL.md b/.opencode/skills/rag-workflows/SKILL.md
new file mode 100644
index 00000000..19339c82
--- /dev/null
+++ b/.opencode/skills/rag-workflows/SKILL.md
@@ -0,0 +1,84 @@
+---
+name: rag-workflows
+description: Retrieval-Augmented Generation (RAG) workflows for document-based Q&A
+---
+
+<!-- //review-2026-02-15 @twishapatel12 -->
+
+## Overview
+
+This skill describes patterns and workflows for **Retrieval-Augmented Generation (RAG)** in OpenWork, with a focus on **local document question answering**.
+
+RAG combines document retrieval with large language models to answer questions grounded in external context rather than relying solely on model knowledge.
+
+This skill is intended to help users design and reason about RAG-style workflows within OpenWork.
+
+## Local RAG with Ollama
+
+A common use case is running RAG **fully locally** using Ollama as the LLM backend. This enables:
+
+- Querying private or sensitive documents  
+- Offline experimentation  
+- Avoiding external API dependencies  
+
+Typical steps in a local RAG workflow include:
+1. Preparing a set of local documents  
+2. Retrieving relevant chunks based on a query  
+3. Providing retrieved context to a local LLM via Ollama  
+4. Generating an answer grounded in the retrieved context  
+
+## Setup (Ollama)
+
+To use the included example workflow, you will need Ollama installed and running locally.
+
+### 1. Install Ollama
+
+Follow the official installation instructions for your platform:
+https://ollama.com/download
+
+### 2. Start Ollama
+
+After installation, ensure the Ollama service is running. On most systems, this can be done by simply running:
+
+```bash
+ollama serve
+````
+
+(or by launching the Ollama app if you installed the desktop version).
+
+### 3. Pull the required model
+
+The example workflow uses **llama3**. Install it with:
+
+```bash
+ollama pull llama3
+```
+
+You can verify it is available by running:
+
+```bash
+ollama list
+```
+
+## Included Example
+
+This skill includes a minimal example workflow:
+
+* **local-ollama-doc-qa**
+
+It demonstrates:
+
+* Retrieving context from local documents
+* Answering questions using a local Ollama model
+
+## Example Use Cases
+
+* Question answering over local markdown or text files
+* Exploring private knowledge bases
+* Prototyping RAG pipelines before production deployment
+
+## Notes
+
+* This skill is **optional** and **not enabled by default**
+* It becomes available via the Skills settings when present
+* This skill currently provides conceptual guidance, patterns, and a minimal runnable example
diff --git a/.opencode/skills/rag-workflows/prompts/doc-qa.md b/.opencode/skills/rag-workflows/prompts/doc-qa.md
new file mode 100644
index 00000000..c1979e06
--- /dev/null
+++ b/.opencode/skills/rag-workflows/prompts/doc-qa.md
@@ -0,0 +1,12 @@
+You are a helpful assistant answering questions based only on the provided context.
+
+If the answer cannot be found in the context, say:
+"I don't know based on the provided documents."
+
+Context:
+{{context}}
+
+Question:
+{{question}}
+
+Answer:
diff --git a/.opencode/skills/rag-workflows/workflows/local-ollama-doc-qa.yaml b/.opencode/skills/rag-workflows/workflows/local-ollama-doc-qa.yaml
new file mode 100644
index 00000000..aba74a96
--- /dev/null
+++ b/.opencode/skills/rag-workflows/workflows/local-ollama-doc-qa.yaml
@@ -0,0 +1,31 @@
+# //review-2026-02-15 @twishapatel12
+
+name: local-ollama-doc-qa
+description: Local document Q&A using Ollama
+
+inputs:
+  documents_path:
+    type: string
+    description: Path to a folder containing local documents
+  question:
+    type: string
+    description: Question to ask about the documents
+
+steps:
+  - id: retrieve
+    type: rag.retrieve
+    with:
+      path: "{{documents_path}}"
+
+  - id: answer
+    type: llm.generate
+    with:
+      provider: ollama
+      model: llama3
+      prompt: ../prompts/doc-qa.md
+      context: "{{steps.retrieve.context}}"
+      question: "{{question}}"
+
+outputs:
+  answer:
+    value: "{{steps.answer.text}}"