Post

Cool project! I caught most of the demo at FoC the other day, happy to stumble across your bluesky :)

You can verify any task on the atelier.dev kanban board and it will automatically build a walkthrough of the acceptance criteria grounded in deep links back into the code that it wrote.

Rather than reading diffs line-by-line, you step through the agent's account of what it did and diff it against your own mental model. This is exactly the sort of high-level decision point which surfaces misunderstandings and gaps in the design. narphorium.com/blog/decisio...

Atelier walkthroughs can use the Chrome MCP in Claude Code to build UI walkthroughs with live screenshots of the app. Grounding the walkthrough in execution exposes gaps in the implementation that you wouldn’t catch with code review.