HOME/FAQ
10 questions · detailed answersQuestions peoplekeep asking us.
The ten we hear most. For billing and subscription questions see the pricing FAQ. For anything else, ping us on Discord or file an issue on GitHub.
Frequently asked.
Tap a question to expand its answer.
01What is Company Agents?
The desktop app for running a company where AI agents and human teammates share the same workspace, board, workflows, and budget. Agents do the work. You and your humans own the decisions that matter.02How is this different from using Claude Code or Codex directly?
Single-agent tools run one task at a time with no structure, no budget, no audit, no delegation, and no integrations. Company Agents is the layer on top of those tools. Keep your Claude, keep your Codex. Company Agents gives them a company to work inside.03What can agents actually do?
With integrations installed, your agents can send emails (Resend), process payments (Stripe), buy domains (Namecheap), deploy websites (Vercel), update your CRM (HubSpot), message customers on WhatsApp, and more. All inside caps and workflows you define.04How do you prevent runaway costs?
Every agent runs under stacked budgets: company, team, agent, workflow, task, and loop. Every overrun creates an escalation you can review. Loop detection catches stuck agents before they burn budget. Built by people who lost $706 in a day to a different tool's runaway loop.05What is the Executive Assistant?
A dedicated AI in your sidebar that works across every company you run. Create workflows, write tasks, configure agents, analyze costs, install integrations, delegate work to your agents. Just ask. It has 12 system-level tools plus every tool your agents have, and it acts immediately.06Can agents delegate to other agents?
Yes. Every agent has a rank and a reporting line. Managers delegate to their reports; reports file updates back up. The full delegation tree is visible in your org chart and audit log.07How do agents get better over time?
Every agent has memory: a private wiki it reads before every run and writes to after. Memory exists at four levels (agent, project, client, and company), so lessons that prove themselves automatically promote to scopes that benefit every future run. Your hundredth website build is unrecognizable from your first.08How do routines work?
A routine is a recurring task with a trigger: cron schedule, signed webhook URL, or API call. Nightly reports, weekly briefings, inbound lead triage, invoice follow-ups. Full run history plus concurrency and catch-up policies.09Does it work offline?
Agents run on your machine, so they keep working when your internet drops. Cloud sync pauses and resumes when you reconnect. Your subscription has an offline grace period for long flights.10Which LLMs does it support?
Any that your CLI adapters support. Bring your Claude, Gemini, GPT, DeepSeek, or local model keys. Company Agents is the layer on top of the CLIs you already use.
Start building yourAI-native company.
Free to start. Your keys. Your data. Your decisions. Every agent runs on your machine.
- Keep your Claude Code, Codex, Gemini CLI, Cursor, OpenClaw.
- Stacked budgets. Loop detection. No runaway costs.
- Free tier forever. Pro at $20/mo. Teams from $15/seat.
~ / company-agents · build status
Still in build
Company Agents
We are still signing the first binaries. Mac, Windows, and Linux releases are next on the roadmap. Follow progress through the changelog.
See what we are shippingHow it worksFree tier plannedNotarized at launch