This static page replays existing synthetic results. It does not run an agent or simulator.
Download the runnable environment and recorded evaluations.
The film and screenshots are the reviewed release artifacts. Case panels contain allowlisted fields from the recorded main and missing-confirmation runs: outcome metrics, check results and selected tool arguments. Prompts, raw events, usage records and tool-result payloads are not included. The displayed model identifier is the recording's configuration, not an independent routing attestation.
Proposed GTS workflows are not implemented integrations and do not imply GTS participation. These authored cases are not a held-out benchmark.
Inter and Space Grotesk are bundled under their original OFL licenses in assets/fonts/. The viewer uses no GSAP or external JavaScript. Project license: LICENSE.