How tldraw built a spatial harness for coding agents on the canvas
max__drake · x · 2026-10-13
tldraw product engineer Max Drake details the "spatial harness" — the systems that let coding agents understand and act in 2D canvas space instead of the linear space of chat.
Key components:
- Transforming the canvas into a representation agents can easily understand and access
- Tools agents can use to edit the canvas
- Supporting systems like canvas linting that warn when e.g. two text shapes overlap
Inspired by CodeAct and Code Mode, agents edit the canvas by writing code. The harness powers tldraw offline, a local whiteboard where users and agents collaborate. The post notes even top coding agents struggle with visual/design work, and spatial context — what's visible, where, and off-screen — is the bottleneck.
More from coding & agent
- Instead of compaction: letting Claude Code wipe its own context and resume from checkpoints — Motiv31 · 2026-10-13
- Amp Adds Interactive Views, Letting You Mock and Test UIs Right in Chat — iannuttall · 2026-10-13
- Using Muse Wonder's CUA as a verifier agent for self-improving code generation — EdwardSun0909 · 2026-10-13
- Meta's new Wonder model praised for finding security bug other agents missed — alexandr_wang · 2026-10-13
- Claude effort levels in practice: medium suffices for 80% of tasks — EricBuess · 2026-10-13
- OpenCode 2.0 now runs in Cloudflare Durable Objects via Agents SDK harness — ritakozlov · 2026-10-13