feat: add AI integration with Ollama and ask command

README.md

@@ -99,4 +99,55 @@ Generates file: `knowledge_base/history.md`
 - Always activate the virtual environment before using the tool: `source venv/bin/activate`
 - Never delete a test case — change the state to `Blocked` if no longer applicable
 - Update `roadmap.md` whenever a step is completed
-- The `venv/` folder should not be shared or pushed to Git
+- The `venv/` folder should not be shared or pushed to Git
+
+
+## Changelog
+
+- 2026-06-30 — Initial project setup
+
+## Requirements
+
+> Already listed above, but for AI integration add:
+- [Ollama](https://ollama.com) installed and running
+- Llama 3.2 model pulled locally
+
+## AI Integration Setup
+
+**1. Install Ollama:**
+```bash
+curl -fsSL https://ollama.com/install.sh | sh
+```
+
+**2. Start the Ollama server:**
+```bash
+ollama serve
+```
+
+**3. Pull the Llama 3.2 model:**
+```bash
+ollama pull llama3.2
+```
+
+**4. Install the Python Ollama library:**
+```bash
+pip install ollama
+```
+
+> Note: The `ollama serve` terminal must remain active while using the `ask` command.
+
+## AI Command
+
+### 🤖 Ask the AI agent a question
+```bash
+python agent/cli.py ask "Your question here"
+```
+
+The agent reads the entire Knowledge Base (test cases, roadmap, guidelines) before answering, ensuring responses are always contextualised to your project.
+
+**Examples:**
+```bash
+python agent/cli.py ask "What should I test next?"
+python agent/cli.py ask "What test cases do we have for the Login module?"
+python agent/cli.py ask "What are the team guidelines for automation?"
+```

agent/cli.py

@@ -227,5 +227,49 @@ def export_tests_history():
 
     click.echo(f"\n✅ Test cases history exported to knowledge_base/test_cases_history.md\n")
 
+# Command: ask a question to the AI agent
+@cli.command()
+@click.argument("question")
+def ask(question):
+    """Ask a question to the AI agent based on the Knowledge Base."""
+    import ollama
+
+    # Load knowledge base context
+    with open(TEST_CASES_PATH, "r") as f:
+        test_cases = json.load(f)
+
+    roadmap_path = os.path.join(BASE_DIR, "knowledge_base", "roadmap", "roadmap.md")
+    with open(roadmap_path, "r") as f:
+        roadmap = f.read()
+
+    guidelines_path = os.path.join(BASE_DIR, "knowledge_base", "guidelines", "guidelines.md")
+    with open(guidelines_path, "r") as f:
+        guidelines = f.read()
+
+    # Build context
+    context = f"""
+You are a QA assistant. Answer based only on the following project information:
+
+TEAM ROADMAP:
+{roadmap}
+
+QA GUIDELINES:
+{guidelines}
+
+TEST CASES:
+{json.dumps(test_cases, indent=2)}
+
+Answer the following question: {question}
+"""
+
+    click.echo("\n🤖 Thinking...\n")
+
+    response = ollama.chat(
+        model="llama3.2",
+        messages=[{"role": "user", "content": context}]
+    )
+
+    click.echo(f"{response['message']['content']}\n")
+
 if __name__ == "__main__":
     cli()