@Ryrenz: Found another killer tool: FounderOS. Starting a company solo? It crams communication, customers, and accounts all into…
Summary
FoundersOS(Founder OS)是一个面向单人公司的开源AI驱动指挥中心,把沟通、客户漏斗、社媒增长、财务等集中到一个仪表盘,由多个具名AI代理分别负责各业务板块。v2 迁移到 BusinessOS 技术栈(SvelteKit、Go、PostgreSQL+Redis、Elixir)。
View Cached Full Text
Cached at: 10/11/26, 09:07 AM
🏢 Found another killer tool: FounderOS. Starting a company solo? It crams communication, customers, and accounts all into one dashboard.
It’s already got 1020 stars and 293 forks on GitHub. The author, Bennett Spooner, runs two companies by himself—this whole setup started as something he built for his own use.
The most exhausting part of working solo isn’t the work itself—it’s constantly switching windows: email in one tab, customer follow-ups in another spreadsheet, social media posting in yet another backend, and at month-end, digging through bank statements. Switch too much, and you forget which customer needs following up.
FounderOS spreads all that across a single real-time dashboard: communication, customer funnel, content growth, brand partnerships, and finances each get their own section. Behind each one hangs a named AI agent handling its piece. The connectors are straightforward too—connected means connected; if not set up, it shows as not set up, no pretending everything’s fine. Outbound writes and scheduled tasks are off by default—you have to turn them on yourself.
When you’re running a company alone, what’s scarcest isn’t hustle—it’s a dashboard that lets you see everything at a glance.
GitHub:
Bennettxai/FounderOS
Source: https://github.com/Bennettxai/FounderOS
Founder OS
A personal operating system for a single person company: a command center that runs your business as a set of AI-assisted “departments.”
Built for whoever is holding the whole thing together: a freelancer, a solopreneur, a solo founder, or a small team where one person still runs point. You set the direction, AI agents cover the departments, and any employees you do have plug into the same board.
Founder OS turns the tabs, tools and mental overhead of running that business into one screen: unified comms, a client funnel, social growth, finances, a knowledge graph, and a roster of named AI agents that each own a real job.
Want to build your own, live, with guidance? That’s what the cohort is for: founderos.sh
v2: Founder OS on BusinessOS
Founder OS v1 was a Next.js + SQLite app. v2 keeps the same operator UI and runs it on the BusinessOS stack:
| Layer | v1 | v2 |
|---|---|---|
| Frontend | Next.js 14 (React) | SvelteKit 2 (Svelte 5) |
| Backend | Next.js API routes | Go (Gin) |
| Operational data | SQLite (better-sqlite3) | PostgreSQL + Redis |
| Knowledge / memory | G-Brain | Optimal Engine (Elixir/OTP), bundled |
| Desktop | — | Electron shell (optional) |
The operator views live under /os. Everything else BusinessOS provides
(workspaces, roles, teams, the module desktop, the terminal) comes with it.
The last v1 release is tagged v1-nextjs.
Quick start
Prerequisites (macOS shown):
brew install node@22 go postgresql@17 pgvector redis elixir python
corepack enable && corepack prepare [email protected] --activate
The schema needs the vector extension. Homebrew’s pgvector bottle builds for
its current Postgres releases (17/18), so pair it with postgresql@17; on another
Postgres, build pgvector for it, or point the launcher at a matching install
with BUSINESSOS_PG_BIN=/path/to/pg/bin.
Then:
BUSINESSOS_HEADLESS=1 make dev-local # Postgres, Redis, Optimal Engine, backend, frontend
# (drop BUSINESSOS_HEADLESS=1 to also open the desktop app)
open http://localhost:5273/register # create your local account
make demo [email protected] # Founder OS workspaces + the demo data
open http://localhost:5273/os # the operator console
make demo creates four workspaces (HQ, Vantage, Launchpad Cohort, Personal),
loads the demo’s seeded data into Postgres and its knowledge into the Optimal
Engine. It is idempotent; re-running it refreshes the demo and moves its dates
to today. No credentials are needed to browse.
Ports live in .env.dev; make dev-local-status shows what is running and
make dev-local-stop stops it. See docs/BUSINESSOS.md for
the platform’s own setup notes.
What you’re looking at
| Route | What it is |
|---|---|
/os | Operator console: pulse row, what needs you, operating volume |
/os/comms | Unified inbox: email, Slack, WhatsApp, calendar and recorder lanes |
/os/funnel | Living client-journey flow and acquisition wheel |
/os/workflows | Multi-step tool workflows, as a tree and a builder |
/os/social, /os/content | Growth dashboard; content pipeline, calendar and lead magnets |
/os/brand-deals, /os/finances, /os/trading, /os/adpilot | Deals, money, brokerage monitor, paid-media planner |
/os/agents, /os/chats, /os/tasks, /os/skills | The agent roster, chat hub, task board and skills |
/os/org, /os/blueprint | Org hierarchy (operator → Conductor → pillars → workers) and the system blueprint |
/os/brain, /os/doctor | The knowledge core on the Optimal Engine, and its health |
/os/integrations, /os/usage, /os/analytics | Connections board, token usage, connector numbers |
/os/roadmap, /os/reference, /os/personas | Phases and quarters, the reference model, persona templates |
Navigate with the sidebar or the command palette (Cmd/Ctrl + K; digits 1–9 jump to the first nine views).
Architecture: demo-first, real-ready
The demo looks alive because of seeded data, but every view reads through the
same path a live install uses, and every connector reports its honest state
(connected, not_configured or error), never a fake green light.
frontend/src/routes/(founderos)/os/ SvelteKit routes, one per view
frontend/src/lib/founderos/ views, kit (slab, charts, motion), chrome
desktop/backend-go/internal/founderos/
api/ GET /api/founderos/pages/<view> and actions
pages/ page builders (read Postgres + connectors + the engine)
connectors/ 24 integrations, honest status, writes behind FOUNDEROS_WRITES
agents/ the agent registry; every seeded agent has a real run()
memory/ retrieval over the Optimal Engine
seed/, etl/ the demo data and the loader behind `make demo`
config/founderos/engine-topology.yaml which engine each workspace's memory lives on
optimal-engine/ the bundled Elixir engine
Data lives in founderos_* Postgres tables, scoped by workspace. Knowledge
lives in the Optimal Engine: agents write sources, signals and pending claims;
facts are promotion-gated, so memory stays trustworthy as more agents write to it.
Safety switches. FOUNDEROS_WRITES=1 allows outbound writes (sends, posts,
payments); FOUNDEROS_CRONS=1 allows scheduled agent runs. Both are off unless
you turn them on.
Credentials
Connectors read keys from the process environment, ~/.founderos/.env, or keys
planted through the Connections board. Nothing else on your machine is read.
Never commit keys. .env* files are gitignored.
Tests
make test # Go + frontend suites
cd frontend && npx vitest run src/lib/founderos
cd desktop/backend-go && go test ./internal/founderos/...
Go tests create throwaway databases on this checkout’s dev Postgres (the
POSTGRES_PORT in .env.dev); set FOUNDEROS_PG_ADMIN_URL to use another.
Write the failing test first.
Credits
Founder OS v2 is built on BusinessOS and Optimal Engine by Roberto H. Luna and the Miosa team. If you use this stack, star their repos.
License
MIT, see LICENSE.
Similar Articles
@GitHub_Daily: Startup founders are often bogged down with miscellaneous tasks like writing business plans, revising contracts, sending emails, etc., leaving them with insufficient time. The Startup Founder Skills project organizes 50 repetitive founder tasks into AI skill packs covering funding, business development, product planning, sales...
Startup Founder Skills is an open-source project that organizes 50 common repetitive tasks for founders into AI skill packs, covering fundraising, sales, product planning, and more. It can be used with multiple AI agents.
@10xmylife: https://x.com/10xmylife/status/2078690248645910965
Independent developer Haiming shared his experience and thoughts on running a one-person company (OPC) in the AI era, including using AI tools such as Cursor and ChatGPT to reduce production costs and improve efficiency, while emphasizing that market risks still exist, and providing an execution methodology.
@mikenevermiss: this is f**king insane. Jack Dorsey just open sourced a free tool for running a one-person business. the GitHub repo al…
Jack Dorsey open sourced a free tool for running a one-person business, combining chat, search, Git, and automation with customizable AI agents. The GitHub repo has 14.4K stars.
@CycleDecoded: Have you ever had this thought: want to start a side hustle or do independent development, but you're already exhausted after work every day and have no energy? I created an open-source project that lets your computer work for you 24/7, cross-platform support for Windows/Linux/macOS: Auto-Company: Fully automatic 24/7 digital…
Auto-Company is an open-source project that creates a fully autonomous AI company running 24/7 using multiple AI agents to ideate, code, deploy, and market products without human intervention.
@interjc: Cloudflare OS looks good, tying together enterprise resources and tasks
Cloudflare OS is an open-source platform that lets everyone in a company build applications, automate work, and securely access internal systems.