Skip to content

otomata-tech/oto

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 

Repository files navigation

oto

Oto is a B2B automation toolkit for AI agents: a platform of connectors and processes exposed both over MCP (Model Context Protocol) and REST, multi-tenant and multi-LLM. You wire your tools (company data, CRM, email, messaging, knowledge base…) once, and any agent (Claude and others) can use them.

This repository is the platform's public entry point. It does not contain the product itself but its architecture doctrine: the structuring decisions (ADRs) and an overview. The code lives in dedicated repositories (see Repositories).

Overview

docs/architecture.md — the diagram and the narrative: a central backend with two faces (MCP + REST), an encrypted credential vault, connectors to third-party systems, and adapters that project a single declaration onto both surfaces. The backend is the center; no façade holds a secret.

Repositories

Oto is not a monorepo — it is a constellation of independently versioned repos. The reusable substrate and the product core are open source; some repos stay private.

Open-source building blocks

The reusable core of the platform is published independently and can be consumed on its own:

Repo Role
oto-core Connector library (oto.tools namespace + config/secrets). Single source of the connector clients, consumed by both the CLI and the backend.
oto-cli oto — Typer CLI façade over oto-core (humans and local agents).
oto-plugin Claude Code plugin: auto-configured MCP connector + a universal skill. Third-party onboarding entry point.
france-opendata French public-data clients (Recherche Entreprises, SIRENE, INPI, BODACC, DVF) + SIRENE stock over DuckDB/parquet.
otomata-calllog Normalized MCP tool-call journal — FastMCP middleware + pluggable sinks, canonical tool_calls schema shared across every Otomata MCP server.
o-browser Browser-automation client (Patchright + remote CDP sessions).
scout White-label Vue 3 + Fastify shell for lead-enrichment platforms (mission-driven theming, generic OAuth/cache/Logto helpers).
memento Structured, sourced, living knowledge substrate for AI agents over MCP (Apache-2.0, hosted at mento.cc).

Product repositories

The platform itself. The core product — backend and dashboard — is open source; the rest stays private. Listed here so the architecture maps onto real repos.

Repo Visibility Role
oto-backend public The backend — encrypted credential vault, orgs, doctrine, monitoring. Two faces: MCP (mcp.oto.ninja/mcp) + REST (/api/*). Imports oto.tools from oto-core. The center of the architecture.
oto-dashboard public Product dashboard for the backend (Vue 3 + shadcn-vue + Tailwind, Logto PKCE). Has no server of its own — the backend is oto-backend.
oto-websites private Sites monorepo: marketing (oto.ninja), scout vitrine, otomata.tech / oto.zone / mento.cc, the Oto Companion extension, and the @otomata/ui design system.
academy private Academy product (change-management / AI adoption) — public best-practices knowledge base + gated accompaniment. Live at academy.otomata.tech.
client bridges private Reference remote connectors (bridge pattern): per-client back-office services that hold the client credential off-platform — see architecture.md.
otomata-private private Working meta-repo — cross-project index, issue tracker, and the ADR drafting copy mirrored into this public repo.

Architecture decisions

The platform is shaped by a series of architecture decisions (ADRs). The enduring ones are summarized in docs/architecture.md; the full decision records are kept internally.

Learn more

  • Platform: oto.ninja
  • Open-source building blocks: oto.ninja/oss

About

Oto — doctrine d'architecture de la plateforme (ADRs + vue d'ensemble). Repo d'entrée public.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors