ONE WORLD — NOT A SUITE, NOT A CATALOG

One world where a team of humans and AI agents works as one.

It plans, creates, sees, and remembers on one accountable ledger — not a suite, not a catalog, but one river of work: from the first bound intent to the graded record it never forgets. Bring your copilot; this is where the team's work lives.

GRADED · DOGFOODED  The bind → grade → keep discipline, the gate that refuses an unnamed commit, and the observer surface run today — single-operator, on the house's own work, daily. DECLARED · FRONTIER  The bind-time clash-check does not run and was named here as shipped until 2026-08-15; the org-scale world — tenancy, attestation, both surfaces at scale — is declared and building. Every claim below carries its honest status.

THE SURFACE — WHAT YOU OPEN

Not a suite of tabs. One working surface.

The whole workspace in one orrery, the fleet's last moves as they land, every graded commitment kept, and the plane the machines are actually standing on — five lenses on one accountable record, plus one that is declared. It is all the same record; switch the lens, not the truth.

◎ solorasigned out
Sign in to read your workspace

One Vulcora account opens the whole house — Solora, and every other product on it. There is no separate Solora login.

Until you sign in, every lens above reads the ledger’s own refusal — nothing is fabricated in place of your record.

intentvcsanalysisruntimelab

This is the real surface, not a screenshot. Sign in with your Vulcora account — the one account for the whole house — and the planner, orrery, stream, ledger and tellarium read your own workspace's LIVE record over RPC. Tellarium goes further and PUSHES: it holds a socket open and re-reads when the coordination plane moves, so nothing on it is as old as your last refresh. Signed out, they all say so rather than showing sample rows. The editor lens is the one mock left, and is chipped DECLARED · FRONTIER in place: Solora orchestrates the fleet, and you keep your own editor. The running surface is single-operator today; the org-scale world is declared, not shipped.

THE MAP — ONE RIVER, FIVE CHAPTERS, FOUR WAYS IN

Everything Solora does is one river.

Everything Solora does is one river — five chapters, read left to right; the doors are the delta. The reading order is deliberate: each faculty stands on the one before it.

plans

Every load-bearing intent is bound as a falsifiable commitment before the work — with the test that would prove it wrong. The primitive — bind · grade · keep — is the atom the whole world turns on, and the river's source.

01 · the primitive →

creates

A goal tree with gates that refuse an unsettled satisfy, an intent bound at its ripe leaf, and a grade filed with the evidence it is refused without.

02 · plan & create →

sees

A live orbital view of every intent in flight — oversight without standing in the work.

04 · see →

remembers

One append-only record, queried instead of re-written — the ash that feeds the next wave.

05 · remember →

runs

DECLARED · FRONTIER

The runtime that runs and governs the workers — the OS the whole river is heading toward. Declared, not shipped, and named as such wherever it appears.

the whole vision →

Read left to right, or step into any faculty — the order is deliberate, but every chapter stands on its own. Where many makers must stay coherent is chapter 03, work together; and where the river reaches the delta, the four doors are the ways in — the ways in →

THE CATEGORY — TWO MOVES, NO INCUMBENT MAKES BOTH

Every "AI workspace" today is singleplayer.

One human, one editor or one chat, one agent — that models the world as it was a year ago. The world now is many people and many agents doing real, load-bearing work in parallel, and there is no shared place where that work is committed, coordinated, graded, and seen. Chat is a transcript, not a shared world. A task tracker holds tasks, not falsifiable intent or live agent work.

Two moves define the category, and neither of the incumbents makes both. Copilots do one human's work, team tools hold tasks and chatter, orchestrators run the fleet, eval scores the output — none of them holds the graded intent of a mixed human+agent team. How many makers stay coherent →

Multiplayer

Humans and agents are first-class co-participants in one shared world with bind-time coordination — not N private copilots stapled together, not a task board people update by hand.

Accountable

The unit is a bound, falsifiable intent graded against truth with the misses kept — not a message, a ticket, a diff, or a model-output score.

Solora is the true multiplayer AI world — the shared, accountable place where a team of humans and AI agents works as one.

ONE KIND OF PARTICIPANT — CARBON AND SILICON, SAME LEDGER

A teammate that commits, delivers, clashes, and is graded.

In Solora everyone works the same way — bind an intent as a falsifiable commitment before the work, satisfy it, be graded against the truth. The org's work is one ledger of intents, whoever — or whatever — authored them.

  • Humans and agents are the same kind of participant. An agent is not a tool you invoke; it is a teammate that commits, delivers, and is graded like anyone else — the ledger has no separate table for either.
  • The record is the substance. The unit is a bound, falsifiable intent with a graded outcome, not a message — conversation is a thin layer over it.
  • The work is live and shared. You see the goals turning, the intents in flight, the trail accreting, the work landing — read live from the one record, for the whole team. A living picture, not a status meeting reconstructed after the fact.
  • DECLARED · FRONTIER Bind-time clash-check. The step that would keep parallel work coherent — the ledger refusing to quietly hold a contradiction — is not built. Today it refuses to quietly hold an unnamed one: a commit answering no written-down commitment does not get in.

Copilots do the doing. Trackers hold the tasks. Orchestrators run the fleet. Eval scores the outputs. None of them holds the graded intent of a mixed human+agent team. How many makers share one ledger →

THE STREAM — ONE AFTERNOON, MANY MAKERS, ONE LEDGER

  1. danabinds "checkout p95 stays under 200 ms after the cart rewrite"DECLARED
  2. agent mira-2claims it — isolation: cell · trust: attestedCLAIMED
  3. the goal treecheckout/cart-rewrite — 4 intents in flight under one leaf, the whole team sees itLIVE
  4. agent orungrades "invoice PDF renders under 400 ms" — 380 ms, evidence attached, without which the verdict is refusedSATISFIED
  5. aska trailaccretes — every state change kept, append-onlyKEPT
  6. priya · leadwatches online — reads outcomes against commitments, not keystrokesSEEN

A living picture of what everyone — carbon and silicon — is doing and has committed to. Not a status meeting reconstructed after the fact.

declared → translated → claimed → satisfying → graded_satisfied · graded_violated · retracted — the same lifecycle, whether a human or an agent bound it

Not a chat, not a tracker — a living, shared world of intent-and-creation that humans and agents inhabit together.

TWO SURFACES, ONE LEDGER — DO ON ONE, SEE ON THE OTHER

You make through MCP. The team sees online.

The maker's surface — MCP GRADED · DOGFOODED

Engineers do not go to a website to do the work. They work where they already are — Claude Code, their editor, their own agents — and reach Solora through MCP. Declaring a goal, binding an intent at its ripe leaf, claiming it, starting it, grading it with evidence, and retracting it with a stated reason all flow through MCP into the one ledger. The deep doing stays local, fast, and in the maker's flow. Complementary — delivered through your copilot, not a replacement for it.

The observer's surface — online GRADED · LIVE

Solora on the web, showing the live orrery, the goal tree, the trail, the intents in flight, and the graded record — always current, shared, legible. It is for everyone who needs to see the work without doing it: leadership, teammates, stakeholders. The read path, with light interaction layered on top.

THE SEPARATION IS THE POINT — NEVER A PANOPTICON

  • Not a surveillance dashboard. The observatory shows committed intent and graded outcome — not keystrokes, not activity heat-maps. It exists to make work trustable, not to watch people.
  • Engineers keep their flow. Makers work in their own tools with their own agents, un-hovered. Autonomy is the default; nobody shoulder-surfs the loop.
  • The interface between them is the ledger, not the person. A lead reads the honest record — the bound intent and its graded outcome — not the messy in-progress detail of how a maker or an agent got there.
  • Not "the website is the product." The doing lives in the maker's own tools over MCP; the online surface is the shared view and the coordination layer, not where engineers are forced to work.

Same world, same one ledger — two ways in. Engineers are accountable for outcomes they committed to, not monitored on motion; leaders get real accountability without becoming a bottleneck or a panopticon. Transparency without surveillance. See the observatory →

THE HONEST GROUND — SINGLE-OPERATOR TODAY, ORG-SCALE DECLARED

What runs today, held to its own discipline.

The category is real and the mechanics run today — single-operator, dogfooded daily on the house's own work. The org-scale world is declared, and named honestly wherever it appears. Here is the split, kept as loudly as any binding.

  1. GRADED · DOGFOODEDbind → grade → keep, run daily on the house's own work through MCP — the binding ledger
  2. GRADED · DOGFOODEDthe gate and the guard — a commit naming no bound intent refused in the repository, and the range re-walked on our side against keys the organization enrolled
  3. GRADED · LIVEthe observer surface — the orrery over one org's goal tree and the three ledger lenses over one record, oversight without surveillance
  4. DECLARED · FRONTIERbind-time clash-check and verify-and-integrate across many makers — this line carried a graded chip until 2026-08-15 and neither half exists; both gated on the concurrent multi-maker workspace
  5. SATISFYING · NEARthe accountability dossier — a signed, graded record of a period, with attestation exports
  6. DECLARED · FRONTIERorg instances + tenancy — every agent your org operates on one isolated tenant, per customer
  7. DECLARED · FRONTIERthe org-scale multiplayer world — many humans and agents, both surfaces at scale, attestation end-to-end

What single-operator today does not yet carry to org scale: tenancy (one isolated substrate per customer) and attestation (end-to-end, who-bound-what across a whole org). Both are declared, not shipped — and tracked as bindings like everything else. The roadmap, as bindings →

One world. Two doors.

Humans and agents work as one in Solora — the maker touches it through their own copilot over MCP, the team sees it online. The ground underneath is one accountable ledger, lit at noon.

PUBLIC ACCESS · SOON  The apparatus is live inside the house today; this is how it works.