Towards
shared minds

AI-native collaboration is here. Turn the traces of your team's work into persistent collective cognition.

Organizational cognition

Every session leaves a trace. Decisions, patterns, handoffs structured and ready to build on. Persistent context beyond individual sessions, compounding into institutional knowledge that any member human or AI can work with.

Multi-agent continuity

Human-AI pairs communicating with each other across a rich, shared context space. Handoffs carry not just state, but the reasoning that produced it enabling speed and situational awareness for people, and depth for AI agents picking up the thread.

Deep pattern recognition

Your organization generates more signal than anyone can track. Deep pattern recognition makes the invisible visible revealing how your team actually operates, where it's converging, and where the coordination infrastructure itself can evolve.

Make it multiplayer /invite

One command brings a teammate into your egregore. They get access to shared memory and coordination workflows from their first session.

Claude Code
$ /invite mia@studio.dev

  Inviting mia@studio.dev...

   GitHub org invitation sent
   Memory access granted
   Knowledge graph linked


  Mia gets access to the organization's accumulated
  context from her first session.
   handoff something to get her started

Atomic unit of AI-native coordination /handoff

A handoff doesn't just communicate the state of the work it carries the context graph that generated it. Decisions, trade-offs, open threads, and the reasoning behind them.

Claude Code
$ /handoff

✎ Handoff saved

  What I did:
  Redesigned the dashboard layout, rewired data layer
  Added responsive breakpoints for mobile


  Still open:
   loading state flickers on slow connections
   accessibility audit on new color tokens


pushed to memory · Alex's Claude will have this context

Automate the git workflow /save

Stage, commit, push, open a PR one command. Automatic enforcement of versioning best practices so that collaborative AI-driven development stays streamlined.

Claude Code
$ /save

Saving to dev/alex/dashboard-redesign

  ✓ 4 files staged
  ✓ committed: "redesign dashboard layout + mobile breakpoints"
  ✓ pushed to origin


PR #87 created
  github.com/studio/app/pull/87
   Mia added as reviewer

See the whole board /activity

What happened today. Who worked on what. Open threads, pending handoffs, quest momentum the full picture of your organization's work. A /activity is worth a thousand standups.

Claude Code
$ /activity

+--------------------------------------------------------+
|  STUDIO EGREGORE * ACTIVITY        alex . Apr 09       |
+--------------------------------------------------------+
|                                                        |
|  3 sessions today. Mia shipped the new color           |
|  system. Jordan's migration is ready for review.       |
+--------------------------------------------------------+
|  HANDOFFS                                              |
|                                                        |
|  [1] * Mia -> you: color token review (2h ago)         |
|                                                        |
+--------------------------------------------------------+
|  . SESSIONS                                            |
|                                                        |
|  Today     Alex: dashboard redesign                    |
|  Today     Mia: color system + design tokens           |
|  Today     Jordan: v2 schema migration                 |
|                                                        |
|  # QUESTS (2 active)                                   |
|                                                        |
|  onboarding-flow    12 artifacts . 3d ago              |
|  perf-audit          4 artifacts . 1d ago              |
|                                                        |
+--------------------------------------------------------+
|  What's your focus?                                    |
+--------------------------------------------------------+

Surface what's beneath /deep-reflect

Your team's knowledge base holds more than anyone has read end to end. Deep-reflect traverses it across decisions, handoffs, quests, and the patterns forming between them.

Claude Code
$ /deep-reflect on the architecture disagreement

  Reading 43 artifacts across 3 quests...

  ⚫️ Spawning CandidateSelection agent...
    9 artifacts selected · 5 weeks · 27 graph edges

  ⚫️ Spawning 3 DeepAnalysis agents in parallel...
    ├─ analysis-1 · 9 tool uses · 22.4k tokens  Done
    ├─ analysis-2 · 7 tool uses · 18.1k tokens  Done
    └─ analysis-3 · 5 tool uses · 20.6k tokens  Done

  ⚫️ 3/3 consensus on core finding.


  You're not having a technical disagreement.
  You're having a disagreement about time.

  Both sides cite the same handoff, the bitter lesson,
  "general methods that leverage computation win in the
  long run", as justification for opposite architectures.
  They agree on the principle. They disagree on when
  "the long run" starts.


   tension· architecture ↔️ architecture
    Same source cited for opposite conclusions.
    The hidden variable is time horizon, not method.

   pattern· 4 "alignment" sessions
    Apparent convergence that isn't. Handoffs
    diverge after every recorded agreement.

   gap· unnamed variable
    No artifact marks when "the long run"
    transitions to "now."

From the lab

Context gardening

Most teams treat knowledge like a warehouse. Store it, label it, forget where you put it. Context gardening is different: it treats organizational knowledge as something alive that grows, connects, and compounds when you tend to it.