Add AI module with A2A wrapper and task agent

- Disable FuzzForge MCP connection (no Prefect backend)
- Add a2a_wrapper module for programmatic A2A agent tasks
- Add task_agent (LiteLLM A2A agent) on port 10900
- Create volumes/env/ for centralized Docker config
- Update docker-compose.yml with task-agent service
- Remove workflow_automation_skill from agent card

docker-compose.yml

@@ -448,6 +448,27 @@ services:
       timeout: 10s
       retries: 3
 
+  # ============================================================================
+  # Task Agent - A2A LiteLLM Agent
+  # ============================================================================
+  task-agent:
+    build:
+      context: ./ai/agents/task_agent
+      dockerfile: Dockerfile
+    container_name: fuzzforge-task-agent
+    ports:
+      - "10900:8000"
+    env_file:
+      - ./volumes/env/.env
+    environment:
+      - PORT=8000
+      - PYTHONUNBUFFERED=1
+    volumes:
+      - ./volumes/env:/app/config:ro
+    networks:
+      - fuzzforge-network
+    restart: unless-stopped
+
   # ============================================================================
   # Vertical Worker: OSS-Fuzz Campaigns
   # ============================================================================