The Smith blog
Guides for building AI agent teams, plus notes on AI employees, agents in Slack and email, and running autonomous agents safely.
Guides
Step-by-step setups for agent teams.
Building an agent team: multiple Smith agents that hand work to each other
How to split work across multiple Smith agents: one role each, three team patterns, and handoffs in Slack and email with people approving key steps.
Hiring an agent: write the role once
Define what a Smith agent is for when you hire it so daily tasks can be one line: what goes in a role, a template, and good versus vague roles.
Setting up accounts for your agents
Decide which account each AI agent signs in with, delegated or agent-owned, set it up, and know what to do when a login needs a person.
Research, then act: a two-agent pattern for follow-up work
Split research and follow-up between two Smith agents with your approval between them, so the agent acting from your account has the narrowest job.
How to set up a recruiting agent team in Smith
Set up three Smith agents that source candidates on LinkedIn and GitHub, screen against your brief and run outreach, with you approving the shortlist.
Customer support with Smith: an intake, resolver, checker and insights team
Set up Smith agents that read your support inbox, look up customers in Stripe and your admin panel, draft and check replies, with people approving sends.
Engineering with Smith: an ops, triage and research team
Set up Smith agents with their own GitHub and read-only dashboard access to watch systems, triage issues and research libraries, handing off in Slack.
Forward-deployed engineering with Smith: one agent per customer
Run one Smith agent inside each customer's tools with accounts the customer provisions, plus shared prep and reporting agents, with an auditable record.
Venture capital with Smith: a sourcing, diligence and intro team
Set up Smith agents that scout deals from gated databases and university research, run diligence with specialists, and find warm intro paths.
Articles
Shadow AI agents: what happens when agents use employee logins
Surveys from SailPoint and the Cloud Security Alliance show many companies can't tell agent activity from human activity. Why that happens, and what to do.
What is a computer use agent?
A computer use agent works a computer like a person: it reads the screen, clicks and types. How it differs from API tools, and why logins are the hard part.
Delegated vs. agent-owned identity: whose account should an AI agent use?
When an AI agent needs to log in, it can use your account or its own. How to choose, and how to keep a record of who did what either way.
What is an AI employee? How it differs from a chatbot
An AI employee is a long-running agent with its own identity, inbox and workspace that takes assignments and works until done. Not a chatbot.
Give an AI agent its own Slack handle and email inbox
Why AI agents work better as teammates in Slack and email than as another app, and how threads, mentions, and replies should behave.
Keeping AI agents safe: sandboxes, scoped credentials, and an audit trail
A practical checklist for running autonomous AI agents with real accounts: isolation, least-privilege tokens, a credential vault, human handoff, and logging.