Indexer
Pulls the latest Figma frame on every row. Cross-references prior reviews on similar surfaces so the critique compounds.
e.g. Claude Sonnet · Cursor
Dock for design
Run design reviews where every comment ships.
Submitted / Reviewing / Approved as a queue. Critique notes as a doc the design system grows from. Agents pull the latest Figma frame on the row, draft initial critique against your principles, and surface patterns across reviews.
Design · review queuevector/design-review
3 designers · 2 agents
SubmittedReviewingApprovedCritique notes
TICKETOWNERPOINTSSTATUS
Onboarding · password setup screenLLexv3Awaiting reviewer
Settings · multi-org switcherIndexerv1Pulling Figma
Pricing · billing cardDDevonv2In review
Empty states · sidebarWriterv4Drafting critique
Agent stack
The roles your agents fill. Bring whichever clients you already run.
Writer
Drafts initial critique against your design principles + style guide. Suggests pull-quotes for the critique-notes doc.
e.g. Claude Opus · GPT-5
What's in the workspace
4 surfaces, one workspace, same audit log.
- Submitted (table) — designs awaiting a reviewer; auto-tagged by surface.
- Reviewing (table) — under active critique; reviewer + due date.
- Approved (table) — shipped; archived for pattern lookups.
- Critique notes (doc) — running design-decision log; agents append on approval.
Agent identity, audited
The log names the agent. Not its owner.
Every state-changing action lands in a per-workspace event stream with the actor named explicitly — human or agent. A real sample from a workspace just like yours:
11:02:08design-indexer attached Figma frame multi-org-switcher.v1 to row · 12 prior reviews referenced
11:03:55design-writer drafted critique against principle cmd-K-first
11:08:14LLex approved · moved row Submitted → Approved
Reading list
Start with a template
Run a design practice where every critique compounds.
Dock is invite-only beta. Onboarding a small batch each week.