What is type.com?
type.com runs your team's automations and agents. You describe the work, and type.com writes the code and instructions, runs them with Claude or Codex on your team's shared connections, and records every run. Teams use it for recurring reports, CRM updates, support triage, and other work that should run on its own.
Is type.com a managed agent platform?
Yes. Every type.com automation comes with a managed agent: Claude or Codex, running on type.com's infrastructure with your Space's connections, skills, and memory. type.com runs the triggers, the sandbox, and the run history, so you don't operate servers, queues, or schedulers, and your automation code never handles credentials.
How is type.com different from Claude Managed Agents, Amazon Bedrock AgentCore, or the OpenAI Agents API?
Those platforms give developers a hosted runtime for agents, and your team builds the application around it: connections to your tools, who can use what, and an interface for colleagues. type.com includes those parts. You describe the work, type.com writes and runs the automation with Claude or Codex on your team's shared connections, and your team reviews runs and changes the automation in type.com or works with it from Slack and email. Claude and Codex run side by side, so you aren't tied to one model provider.
How does a type.com automation run?
A trigger starts each run: a schedule, a Slack message, a GitHub or Linear event, a webhook, or an email. The run can combine code for the fixed steps, your team's skills, the managed agent for judgment, and an isolated Linux sandbox for commands and scripts. You test it in a private thread before you enable it, and every run records its steps, output, and errors for your team to review.
Can type.com run Claude Code or Codex agents on a schedule?
Yes. Describe the job, such as “every weekday at 9:00, summarize new support tickets,” and type.com runs it on that schedule, anywhere from every minute to once a month. Runs happen on type.com's infrastructure, using Claude through the Claude Agent SDK (the harness behind Claude Code) or Codex, so no laptop or server has to stay on. Engineers can also write automations in TypeScript and deploy them from Claude Code or Codex with the type.com CLI.
How does type.com handle credentials and permissions for AI agents?
You connect each app once with OAuth, MCP, or an API key. type.com stores the credential and adds it to each call, so nobody pastes keys into prompts or code. Each connection is assigned only to the Spaces that need it, Space membership controls who can use it, and you choose which connected account each scheduled automation acts as, such as a shared team account or your own.
Can people who don't code build automations in type.com?
Yes. Describe an automation in plain language, test it in a private thread, and enable it. Results land where people already work: in type.com, Slack, or email. Engineers can build and manage the same automations from Claude Code, Codex, the type.com CLI, or MCP.
How much does type.com cost?
Plans start at $50 a month for 2 team members and 100 AI usage credits. One credit is $1 of AI usage at the model provider's rates, and every plan starts with a 14-day free trial.
Can we use our own Claude or Codex subscription?
Yes. Team members can connect supported Claude or ChatGPT subscriptions, and eligible requests run through them instead of the workspace's included AI usage.
Is type.com the same as type.ai?
No. type.com runs shared automations and agents for teams. type.ai is a separate AI writing editor for books and stories, with its own accounts and plans.