Run a one-person company on five building blocks: a browser agent, a coding agent, scheduled routines, memory notes and approval gates
A solo founder of a physical-product business with no employees describes the five pieces of an agent setup and the rules that keep it safe: run a workflow by hand twice, write memory notes, gate money paths, then schedule it.
Evidence: The author reports this. We have not checked it beyond reading the source.
The business problem
Most small-business work sits behind logins with no usable API, and unattended agents can fail quietly or take irreversible actions.
What was tried
Use a browser-operating agent to reach seller dashboards, ad platforms, email and government sites with your own logins, and a coding agent that reads a repository, writes changes, runs a review pass and deploys. Schedule recurring routines, such as a 7 a.m. scan of grant portals or a daily draft of posts and replies, and require each to report 'nothing changed' explicitly so that silence signals a fault. Keep structured memory notes per site, project and decision rather than one large file. Make agents draft and you approve, with explicit sign-off for payments, filings, public posts, customer emails and production deploys. To start: pick the workflow that wastes the most time, run it manually with an agent twice, write the notes, gate the money steps, and only then schedule it.
What was reported (mixed)
The author says the setup covers the routine work of roughly two to four roles at a tool cost of a few hundred dollars a month. No revenue, customer or hours-saved figures are given. He lists failures: a deploy that shipped wrong discount tiers, a refund reported as done that the gateway had blocked, and a scraper that returned stale data for nine days.
Limitations
Written by an unnamed founder who sells a paid playbook ($19) and runs an affiliate programme, so the piece promotes that product. The article names no specific tools. The author says agents fail regularly and gates make failures survivable but do not make them reliable, that memory degrades if it grows into one file, and that scheduled jobs can produce confident errors if never validated by hand. Agents do not replace the founder's judgment on pricing, brand, product or legal strategy.
What you need
A browser-capable agent, a coding agent, somewhere to run scheduled jobs, and a habit of keeping written notes for the agents.
Sources
- Solo Operator Playbook (seller of the playbook; discussed on Hacker News) ↗ Vendor case study, published July 28, 2026
Source published: July 28, 2026. Last reviewed here: October 14, 2026. Spot a mistake? Tell us.
Related workflows
Plan ten Claude Code skills for a small service business, one a day, and review the first 20 to 50 runs before anything is sent or paid
An agency guide lists ten back-office tasks to turn into Claude Code skills, such as inbox triage, reply drafts, invoice processing and weekly reconciliation, with a rule to review early runs by hand before irreversible actions.
Run a solo consulting practice on Claude Cowork agents with a three-pass quality check before anything reaches a client
A former strategy consultant describes a plugin that gives a solo consultant agents for proposals, kickoff packs, research and deliverables, with source checking and two independent audits before delivery.
Run a one-person business with a fleet of single-job Grok Bots with approval gates
A solopreneur describes a fleet of named Grok Bots, each with one job and explicit anti-jobs, handling briefs, inbox triage, research and social drafts, with approval before anything is published or sent.
Install Hermes Agent for scheduled jobs and a self-built skill library, reachable from Telegram, Slack or WhatsApp
The README for Nous Research's open-source Hermes Agent describes persistent memory, skills it creates after complex tasks, a natural-language scheduler and one gateway for chat apps.