diff --git a/.agent/workflows/deploy_to_production.md b/.agent/workflows/deploy_to_production.md
index 3d52b12..f6f6515 100644
--- a/.agent/workflows/deploy_to_production.md
+++ b/.agent/workflows/deploy_to_production.md
@@ -9,13 +9,14 @@
 
 1. **Automated Secret Fetching**: The `scripts/remote_deploy.sh` script will automatically pull the production password from the GitBucket Secret Vault if the `GITBUCKET_TOKEN` is available in `/app/.env.gitbucket`.
 2. **Sync**: Sync local codebase to `/tmp/cortex-hub/` on the server.
-3. **Proto Regeneration**: If `ai-hub/app/protos/agent.proto` has changed, the agent must regenerate the Python stubs:
-   ```bash
-   cd /app/ai-hub/app/protos && python3 -m grpc_tools.protoc -I. --python_out=. --grpc_python_out=. agent.proto
-   # And for the agent node
-   cd /app/agent-node && python3 -m grpc_tools.protoc -Iprotos --python_out=. --grpc_python_out=. protos/agent.proto
-   ```
-   > **CAVEAT**: Because you changed the protobuf interface, all clients must be updated as well! Remember to rebuild and restart all agent nodes (like `cortex-test-1` and `cortex-test-2`) whenever the `.proto` files change so they have the latest interface stubs.
+3. **Proto & Version Management**: 
+   - If `ai-hub/app/protos/agent.proto` has changed, regenerate stubs:
+     ```bash
+     cd /app/ai-hub/app/protos && python3 -m grpc_tools.protoc -I. --python_out=. --grpc_python_out=. agent.proto
+     cd /app/agent-node && python3 -m grpc_tools.protoc -Iprotos --python_out=. --grpc_python_out=. protos/agent.proto
+     ```
+   - **Triggering Client Updates**: To force all active agents (like the Mac Mini) to pull new code or skills, bump the version in `/app/agent-node/VERSION`.
+   - **Auto-Update**: Once deployed, agents check for version mismatches every 5 minutes and will automatically re-bootstrap themselves if a newer version is detected on the Hub.
 4. **Migrate & Rebuild**: Overwrite production files and run `bash scripts/local_rebuild.sh` on the server.
 5. **Post-Deployment Health Check**: Perform a backend connectivity check (Python Trick). Only use `/frontend_tester` as a last resort if UI behavior is suspect.
 
diff --git a/.agent/workflows/deployment_reference.md b/.agent/workflows/deployment_reference.md
index f8d5926..c11ad5f 100644
--- a/.agent/workflows/deployment_reference.md
+++ b/.agent/workflows/deployment_reference.md
@@ -58,7 +58,6 @@
       3. Triggers Docker Compose with multiple layers:
          - `-f docker-compose.yml` (Generic common config)
          - `-f deployment/jerxie-prod/docker-compose.production.yml` (Jerxie-specific overrides)
-         - `-f deployment/test-nodes/docker-compose.test-nodes.yml` (Optional test nodes)
       4. Rebuilds and starts the containers.
       5. Performs automated database migrations running parallel idempotent logic via the `Uvicorn` startup lifecycle.
 
@@ -74,7 +73,6 @@
 
 *   **`docker-compose.yml`**: Baseline configuration. Uses local Docker volumes and generic localhost endpoints. Default for new users.
 *   **`deployment/jerxie-prod/`**: Contains Jerry's specific production overrides (NFS volume on `192.168.68.90`, SSL/OIDC endpoints for `ai.jerxie.com`).
-*   **`deployment/test-nodes/`**: Contains internal test node definitions for simulation.
 *   **`scripts/`**: Centralized automation scripts for remote sync, local rebuilding, and node registration.
 For any automated AI attempting to debug or push changes:
 * **Primary Source of Truth**: This file (`.agent/workflows/deployment_reference.md`) defines the architecture rules.
@@ -83,6 +81,21 @@
 
 ---
 
+## 5. Agent Node Auto-Update Architecture
+
+Cortex supports zero-touch maintenance for distributed agent nodes (like Mac Minis or remote VMs). 
+
+### How it Works
+1. **Version Monitoring**: The running Agent process (`agent_node/main.py`) runs a background `AutoUpdater` thread.
+2. **Polling**: Every 300 seconds, the agent calls `GET /api/v1/agent/version` on the Hub using the `SECRET_KEY` for authentication.
+3. **Trigger**: If the Hub reports a version higher than the one found in the local `VERSION` file, the agent:
+   - Downloads the latest code bundle (including `skills/`) from `GET /api/v1/agent/download`.
+   - Delegates to `bootstrap_installer.py --update-only` to extract and install new dependencies.
+   - Restarts itself (`os.execv`) to load the new code.
+4. **Resiliency**: The installer uses `--ignore-installed` to prevent common Python/Anaconda metadata errors from stalling updates.
+
+---
+
 ## 4. Reference Troubleshooting & Known Pitfalls
 
 If production encounters a bug or routing fails, these are the historically primary offenders:
@@ -119,4 +132,5 @@
 **Root Cause**: The Envoy FilterChain (Listener SNI Map) doesn't trace back to a correct, valid Docker IP:Port allocation.
 **Verification Check**:
 * Query the Control Plane API: `curl -s http://192.168.68.90:8090/get-cluster?name=_ai_unified_server`.
+* **Explicit Port Matching**: Envoy in production is configured to match both `ai.jerxie.com` and `ai.jerxie.com:443`. If a gRPC client (the Agent) includes the port in its authority header, Envoy must have it in its `virtual_hosts.domains` array or it will throw a 404.
 * Make sure `portValue` in the JSON Endpoint equates to the one published in `docker-compose.yml` (`8002` vs `8000`). If mismatched, you must format a JSON package and `POST` it to `/add-cluster` utilizing the EnvoryControlPlane workflow.