One Graph, Every Lens: How Multi-Board Views Eliminate Board Fragmentation
MetaSpark's multi-board views query the same ranked task graph through different persona lenses. Agents keep each view current overnight without manual sync, no more parallel kanbans drifting out of sync.
The Board Fragmentation Problem
Every modern ops team runs on 5-15 work-tracking SaaS subscriptions. The founder uses Linear. Engineering uses GitHub. Customer ops lives in Slack. The CEO maintains a Notion dashboard. What should be a single source of truth has fragmented into N boards, each with its own copy of the task, each updated on a different schedule, each lying about what's actually happening.
When meeting notes hit Notion, they're stale in Linear within hours. When a PR lands in GitHub, the founder's dashboard doesn't know about it. When a support ticket closes in Slack, nobody remembers to mark it done in the ops kanban. Teams end up with two workflows: the real one (whatever happened in the tool where the work was) and the shadow one (the board someone's supposed to update).
The cost is invisible but real. Engineering spends 90 minutes per week grinding synchronization work. Ops burns time asking people for status updates instead of acting. Founders make decisions based on stale dashboards. Nothing breaks loudly; everything just drifts.
What We Built
MetaSpark v2 ships with Multi-Board Views: the same ranked task graph, surfaced through different persona-specific lenses. The founder gets My Day, a violet-edged Up Next, in-progress rows with green fill, stuck rows pulsing red before deadlines. The engineering lead sees Engineering, PRs ranked by dependency and calendar fit, commits and reviews threaded below. Customer ops sees Customer Ops, tickets ranked by SLA, escalations flagged, resolution paths pre-populated.
Each lens is a query, not a separate board. Switch personas in one tap. The underlying graph stays unified.
The breakthrough is that agents maintain these views on your behalf. After a meeting ends, an agent reads the notes, extracts decisions and action items, routes them to the right person, tags the source ticket, and writes the status back to Linear so the graph updates. The founder's My Day refreshes overnight. The engineering board reflects today's commits before standup. Customer ops sees newly resolved tickets before the support team asks for a sync.
No one touches the boards. They stay current because the harness, the operating layer underneath, is running agents that actually do the work.
Concrete Behavior
Here's what happens in a real workflow:
1. A customer escalation lands in Slack at 3pm. An agent tags it as urgent in the underlying graph. 2. Customer ops' lens immediately shows the ticket at the top, with pre-drafted severity language pulled from your prior escalation responses. 3. The founder's My Day doesn't change, this isn't a founder decision, but the dependency planner flags it against any open contracts this customer's tied to. 4. By morning, the ticket's marked resolved. The agent has written the status back to Slack, updated the linked Linear issue, and cleared it from the graph. 5. Every view refreshes. No one attended a sync meeting. No status update was manually typed anywhere.
The audit log below each view shows every agent action with timestamps and rollback options. You see exactly what changed, when, and why.
Why This Matters
Most agentic systems are single-shot demos: you ask a bot a question, it gives you an answer, and the demo ends. They're impressive on stage. They're useless in practice because they don't touch the systems your team actually relies on, and they don't stay in sync with reality.
MetaSpark's breakthrough isn't a new model or a smarter chat interface. It's the harness. The operating layer underneath that lets any agent, yours, ours, Claude-driven, custom-built, actually get work done against real systems. Multi-Board Views is what that looks like in practice: every person on your team sees their work ranked by what matters, updated by agents running overnight, all grounded in a single shared graph.
This eliminates the synchronization tax. Founders make decisions on current data. Ops moves faster because they're not chasing status updates. Engineering sees blockers before they break the sprint. And no one has to choose between "update the board" and "do the work."
How to Try It
Multi-Board Views ships as part of MetaSpark v2, live now. Connect your Linear, GitHub, Slack, Notion, Gmail, and Calendar. Define your personas in Agents, My Day, Engineering, Customer Ops, whatever fits your team. The harness will rank the unified graph through each lens, and agents will keep those lenses in sync with reality overnight.
If you run on tools we don't have native connectors for, internal Rails CRM, SharePoint, partner APIs, use the Connector Compiler. An agent reads your system's documentation, infers the resource graph, and ships a working integration in under two minutes. Universal reach without a roadmap.
See it live. Read the docs. Try a three-person team first. By week two, you'll stop opening your individual boards and just use MetaSpark.