Director-AI Studio
Present and inspect streaming-AI safety examples, with a path from an educational demonstration to an evidence review.
Phase 0 scaffold · integration under development · Source review: 2026-09-26
What it is for
Present and inspect streaming-AI safety examples, with a path from an educational demonstration to an evidence review.
Who it is for
Developers and reviewers studying contradiction detection, stream halting and the limits of AI safety claims.
Documented capabilities
- A thin federated panel presents manifest information and a link to the fuller portal.
- The planned academy workflow connects learning, demonstration and audit around contradiction halting.
- The source includes a WASM consumer demonstration; outcomes must be assessed with accuracy and false-positive evidence.
A typical workflow
Read the safety model and demonstration assumptions → inspect a supported example → examine its halt decision and evidence → review false-positive and accuracy limits before application.
Scope and evidence
The Studio README labels this Phase 0 and pre-ratification. A panel or consumer demonstration does not establish a completed hosted academy or general hallucination prevention. Build and integration readiness must be checked for a specific deployment.
Local setup and dependencies
Use the studio-web README for the current scaffold and shared SDK dependencies. For working core capabilities, follow the main Director-AI project documentation.
Availability and access
The Director-AI core and the Studio integration have separate readiness and access descriptions. See the project's current licence and access pages; no finished hosted Studio is offered here.
Read the project sources
Project overview · All Studios and their roles · Discuss a research or deployment workflow