We have deployed version 1.6.0 to the adGPT infrastructure. This update significantly upgrades the AI's diagnostic capabilities, unifying Meta and Google performance data into a single source of truth, while simultaneously delivering a massive polish to the Studio's user experience.
All active workspaces have been updated automatically.
1. Omni-Channel Meridian Diagnostics
Calculating true ROAS when running both Google and Meta Ads is notoriously difficult due to overlapping attribution. We have upgraded the Chief Architect AI to solve this autonomously.
- The Hybrid Truth: The monitoring engine now queries your centralized data warehouse directly, analyzing both Meta Ads and Google Ads simultaneously. It uses Meridian's Bayesian ROI calculations as the ultimate "Source of Truth" for triggering growth alerts.
- Cross-Network Auto-Investigation: If the OS detects an anomaly (e.g., a drop in Meta efficiency causing a decrease in Google Search conversions), it will automatically dispatch a diagnostic alert to your HUD.
- Rate-Limit Immunity: By routing analytical queries through your dedicated data warehouse rather than hitting the live Meta Graph API, the OS can run deep, high-frequency forensic audits without triggering platform lockouts.
2. The Omni-Box & Traffic Simulation
We have completely overhauled how operators interact with the Studio canvas and the AI Copilot.
- Simulate Traffic: We have replaced complex developer testing modes with a unified "Simulate Traffic" protocol. You can now safely push dry-run data through your automations to verify complex routing logic before deploying them live.
- The Tactical Omni-Box: The AI Copilot interface has been redesigned into an OLED-black Omni-Box. It features auto-expanding inputs, hardware-accelerated bounding, and premium tactile mechanical feedback when selecting AI-generated blueprints.
- Infinite Loop Protection: We integrated advanced graph-theory depth limiters into the canvas physics. If you accidentally wire a circular automation loop, the OS mathematically detects the infinite loop and safely breaks the connection, preventing browser freezes and wasted compute.
3. Generative Media Node Parity
The multimedia nodes inside the Visual Studio have been upgraded to match the exact capabilities of the main Chat interface.
- In-Node Audio Previews: When building Text-to-Speech automations, the dropdown menus now feature persistent, memory-safe audio preview buttons. You can audition synthetic voices directly within the node properties before spending execution credits.
- Execution Guardrails: We injected strict formatting guardrails into the media nodes. The OS will automatically enforce native aspect ratios (e.g., locking Video Generation to 16:9 or 9:16) and cap Multi-Speaker outputs, ensuring your background automations never fail due to bad API parameters.
- Global Localization: All media generation controls have been fully internationalized, ensuring that multi-modal configurations map perfectly regardless of your agency's operating language.