Benchmark replay
Code Review Arena
Pick a case and reviewer. Run the patch through tests, validators, and repair scoring.
Demos
Twelve working demos tied to the project code.
12 projects
Benchmark replay
Pick a case and reviewer. Run the patch through tests, validators, and repair scoring.
Debugging capture
Record a failing test, apply the edit, rerun it, and generate a PR brief.
Task runner
Run three browser tasks through intent parsing, actions, risk checks, blockers, and completion.
Score explorer
Compare three saved runs. Turn signals on or off to recalculate the index.
Canvas lab
Enter a goal, render the matching fixture, switch views, and add a streaming update.
PII check
Choose a prompt. Run the project’s email, phone, SSN, and API-key rules, then inspect the masked text.
Combat slice
Fire at the Mirror Carapace. Frontal shots are blocked while it is sealed, and rear-core hits deal double damage while it vents.
RAG request path
Follow one research-computing question through the bias check, Azure Search, prompt assembly, response, and validation.
Data query router
Ask a question over a small sales dataset. Get a chart, table, or written answer.
Video workflow
Choose a topic, visual style, and duration. Build the script, voiceover, images, captions, and Remotion preview.
Incident repair
Choose an incident. Run detection, classification, the matching repair handler, SQLite logging, and the alert.
Weather interface
Search a sample city, switch units, and view current conditions with a five-day forecast.