Javis OS
Javis OS is an Agentic AI that assists entrepreneurs in business tasks, helping them grow their work.
AI translation
Books you keep by talking, under rules that refuse a guess. On saybooks.io with Claude, or entirely on your Mac, with an iPhone app to ask from anywhere.
BusinessJavaScript · HTMLGitHub @alvisoOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
No updates detected yet. GitHub releases will appear here.
Owners of public repositories under a personal GitHub account can sign in to verify ownership and manage the introduction themselves. No administrator approval is needed.
Sign in, generate a verification file, and commit it to this repository’s default branch within 30 minutes. Files on other branches or forks do not count. Then check the file below to manage the project without administrator approval. Keep the file to retain your access.
If needed, you will go to sign-in. After signing in, return here and press your chosen verification button again.
Run /nomorevibe verify in your project folder to verify ownership and manage the introduction yourself.
Verify in one minuteAsk about other verification options
Saybooks — open-source books an AI agent can keep: invoicing, order to cash, CRM, job hunt and personal finances over one MCP connector, with rules the agent cannot break and one audit trail for people and agents. AGPL-3.0.
Taken from the repository README
Loading the full README. The saved excerpt is shown below.
The books you can talk to. The books that stay books.
Saybooks is an open-source ERP where your agent and your interface are the same system: one
command registry, one rulebook, one audit trail. Tell Claude to ship the order, issue the
invoice or import a card statement, or click the button: same command, same rules, same record
of who did what and why.Read the full README ↗The books you can talk to. The books that stay books.
Saybooks is an open-source ERP where your agent and your interface are the same system: one command registry, one rulebook, one audit trail. Tell Claude to ship the order, issue the invoice or import a card statement, or click the button: same command, same rules, same record of who did what and why.
Live: saybooks.io. Try the demo with no signup, or sign in with Google
and pick the modules your books carry. Connect Claude by adding https://saybooks.io/mcp as a
connector (OAuth, no keys to paste); the same books are in the browser as ordinary pages.
Watch: Saybooks process demos, five short videos recorded live against the demo books: purchase orders and receiving, inventory value, the month-end close, bank reconciliation, and sales tax.
Fourteen modules today, every one spec-first with executable conformance scenarios:
| module | for | what it keeps |
|---|---|---|
| solo (Invoicing) | freelancers | clients, invoices with a link and a PDF, payments, outstanding; any currency, your tax scheme |
| o2c (Order to cash) | small businesses | quotes, orders, shipments, invoices, receivables, with a credit gate the agent cannot talk past |
| crm | small businesses | campaigns, accounts, contacts with sources, gaps, pipeline; outreach drafts the agent writes and a person sends; events with their source and a calendar |
| prospect | anyone with a bought list | a holding area in front of the CRM: the agent stages and judges rows, only a person promotes |
| jobhunt | people open to work | postings, applications, interviews, recruiters, a duplicate guard, one next action |
| p2p (Purchasing) | small businesses that buy stock | purchase orders at the vendor's price, goods received against them, vendor bills held to a 3-way match before a person approves paying |
| inventory | small businesses that hold stock | every stock move kept, the moving average cost, each shipment's cost and margin, counts with reasons, a value that ties to the books |
| purchases (Personal finances) | anyone | accounts and net worth, transfers between them, card dues, interest and refunds, expected bills, autopay and what is due, work expenses paid back, tax tags and the year; bank and card statements and receipts the agent reads: rows with provenance, review, subscriptions, spend |
| bankrec (Bank) | anyone whose accountant asks "does the bank tie?" | each statement against the books, deposits matched to recorded payments, items in transit named, a sign-off by a person |
| loans (Personal finances, Loans) | anyone with a mortgage or a car loan | the lender's figures, each payment split into principal, interest and escrow, what is owed and when it ends, interest per year, a check against the lender |
| salestax (Sales tax) | anyone who charges sales tax and files it | the return for a month or quarter from what was invoiced and credited, a filing a person records, the payment, a liability that ties |
| gl (Month-end) | anyone closing the books | a close checklist, adjusting entries that balance and can reverse, closed periods that stop moving, corrections in the next open period |
| bridge (Ledger) | anyone with an accountant | the hand-over to QuickBooks, Xero or a plain CSV: map their chart once, export a period, keep what went and through when |
| core | every space | customers, items, the company profile, search, audit, setup |
"Conversational ERP" usually means a chat layer bolted onto a system that doesn't know the agent exists. Saybooks inverts that: every business act is declared once, and everything else is derived from that declaration:
| derived | so that |
|---|---|
| the MCP tool (name, doctrine, JSON Schema, annotations) | the agent sees exactly what the UI can do |
| the UI form (fields, widgets, validation) | the human sees exactly what the agent can do |
| the guard evaluation | a greyed-out button's tooltip, the thrown refusal, and the agent's answer are the same sentence, asserted by test |
| the audit row | every write from either surface lands in one log with an actor and a reason; refusals included |
Nothing outside the registry may call a handler. A click and a tool call go through the same
command, the same guards, the same transaction. The two surfaces cannot drift, because neither
is authored, and test/contract.test.js fails the build if anyone tries.
Agent-first, not parser-first: where a file is involved (a statement, a receipt), the agent reads it and hands over what it read; the module checks the control totals, keeps the provenance of every row, and refuses what does not reconcile. No bank parsers to maintain.
src/registry.js declare once; derive tools, forms, guards, audit; enforce
src/workspace.js a SQLite database per space (per person, per sandbox)
src/members.js capability-token identity: named members with roles
src/oauth.js OAuth 2.0 front door for MCP clients (DCR, client-id metadata documents)
src/mcp-http.js MCP over streamable HTTP; one tool list per space's modules
src/document.js the invoice document: one PDF renderer, previewed as its own pages
src/modules/core, o2c, crm, prospect, jobhunt, solo, purchases, bridge
specs/<area>/spec.md the area spec: calibration, entities, invariants, read models
specs/<area>/scenarios/ executable conformance scenarios (refusals are contract)
src/conformance.js replays scenarios through the real registry; keeps evidence
server.js workbench + hosted demo + MCP + OAuth (all one registry)
mcp-server.js stdio MCP for local development
test/contract.test.js the 24-gate contract (see below)
test/smoke.mjs every workbench view and detail panel rendered in headless Chrome
test/local/ plain-language scripts run against a local model, checked in the database
Access control: every command declares a permission tag; four roles (owner / controller / clerk / viewer); enforcement at the registry choke point; denials are one-sentence refusals, logged. An agent connecting through a member's key is that member's delegate: their name, their permissions. Every write carries a reason; reads are never logged.
Money is integer minor units everywhere, with a currency; sums are per currency and never cross. Nothing is invented: a missing amount, date or name is asked for, never guessed.
Every module — present and future — is held to: MCP/UI parity · namespace prefixes · doctrine on every write · guards declared · table ownership (no cross-module writes) · a tool budget · module mounts · one-sentence rule · audit behavior · fixture replay · spec conformance (an implementation claiming an area must map every act and pass every scenario) · permissions (unpermissioned commands do not ship) · spec version (a module implements the version its acts file declares) · human-only (a field or a whole act a person must do, refused to agents whatever their role) · the door (staged rows are judged by an agent and promoted only by a person) · the claim gate (what a campaign may say is a person's to set, and holds on edit) · bills (only a person approves paying one; a payment is recorded after it happened, against a statement row that matches it).
The endgame is competing implementations of the same area, certified by replaying the same scenario files — the spec speaks in acts, not commands, so any conforming module runs them.
npm install
npm test # the 24 gates
npm run smoke # every workbench view and panel, in headless Chrome
npm run local -- --model <id> --script plain-invoicing --instructions off # a local model, plain prompts
npm run demo # a full quote-to-cash run, human and agent interleaved
npm start # workbench on http://127.0.0.1:8140
npm run mcp # stdio MCP server (OTC_WORKSPACE=you)
Register for a Claude Code session: claude mcp add saybooks-local -e SAYBOOKS_WORKSPACE=$USER -- node <path>/mcp-server.js
Saybooks is a plain MCP server, so a model on your own machine can keep the books with nothing
leaving it. Tested with Gemma 4 26B A4B in LM Studio on a 32 GB laptop: a full invoice cycle and
a full statement cycle from plain sentences, no coaching. In LM Studio's mcp.json:
{ "mcpServers": { "saybooks": {
"command": "/usr/local/bin/node",
"args": ["/path/to/saybooks/mcp-server.js"],
"env": { "SAYBOOKS_WORKSPACE": "main", "SAYBOOKS_MODULES": "core,solo", "SAYBOOKS_ACTOR": "gemma-4" }
} } }
SAYBOOKS_MODULES picks the tools the model sees (core,solo is 35 tools; core,purchases 60;
all fourteen modules is 236 and too many for a small model). Load the model with one concurrent slot.
What that took, and the six fixes it forced in the tools, is in
saybooks.io/notes/local-models; the plain-language
harness that measures it is npm run local.
Any currency per invoice but no conversion between them (sums stay per currency), one company
per space, no tax engine (rates are captured and frozen per line; determination is an
integration's job), no outbound anything: Saybooks never emails a customer, charges a card, or
books a shipment. It records what happened. The ledger hand-over writes import files for
QuickBooks Online and Xero; it does not post over their APIs. Each area's deferred list, with
reasons, is in its spec under specs/.
AGPL-3.0-only. Copyright (C) 2026 Peter Varga. If you run a modified Saybooks as a service, share your changes — that's the deal.