Frequently Asked Questions About Nick Orgo Solo AI Agent Business Playbook
22 answers covering everything from basics to advanced usage.
// Basics
What is the 'agents build agents' concept?
Agents build agents means using an existing AI agent (like a Hermes agent via Orgo MCP or Claude Code) to deploy, install, and configure new client agents inside Orgo cloud VMs. Instead of manually setting up each agent from scratch, you issue natural-language instructions to your setup agent, which handles installation, tool configuration, and skill setup. This is the core fulfillment leverage that makes a one-person business economically viable at scale.
What is Composio and why is it required for every agent?
Composio is a single MCP connector that links AI agents to thousands of apps including Gmail, Slack, Notion, and GitHub. It handles authentication and tool calling, which are the biggest security and setup friction points in agent deployment. Every agent in the playbook gets Composio as a non-negotiable component because it eliminates the need to build custom integrations for each client's tool stack, dramatically reducing setup time.
What is the Executive Abstraction Layer?
The Executive Abstraction Layer is the universal set of pain points shared by decision-makers across all target industries: too many emails, too many meetings, too many follow-ups, too many open loops, and context scattered across too many projects. You lead every sales conversation with these universal pains because they resonate immediately, then layer in vertical-specific solutions. This approach works regardless of industry because every executive experiences these problems.
What is Agent Mail and why does it matter for client perception?
Agent Mail gives each deployed AI agent its own named email address (e.g., mia@clientdomain.com or aria@firmname.com), enabling the agent to send and receive emails as a named persona. This reinforces the digital employee framing — the client and their contacts interact with what feels like a real team member, not a bot. It is a critical perception layer that justifies the $5,000/month premium and makes the agent feel integrated into the client's business rather than bolted on.
What if I don't know how to code — can I still follow this playbook?
The playbook is designed for operators, not developers. Agent deployment uses natural-language instructions via Claude Code, Codex, or an existing Hermes agent — you describe what you want in plain English and the setup agent handles configuration. Composio eliminates custom integration coding. Obsidian is markdown, not code. The key skill is understanding business problems and client communication, not programming. However, basic comfort with command-line terminals and cloud VM concepts will accelerate your first deployment.
// How To
How do I set up an Obsidian vault for a client's AI agent?
Create a markdown-based Obsidian vault pre-loaded with structured context about the client's people (names, roles, preferences), projects (active cases, deals, campaigns), workflows (how they handle emails, approvals, follow-ups), tools (Gmail, Slack, CRM), and key decisions. Sync meeting notes from Granola into the vault automatically. The more organized and detailed the vault, the more the agent feels like it truly knows the business — agents without structured context feel generic and forget everything.
How do I set up watchdog monitoring for client agents?
Instruct each agent to set up a gateway watchdog that auto-detects and auto-restores crashed connections. Configure agents to email you from their Agent Mail address when any cron job, skill, or gateway fails. Hermes agents are more gateway-stable than OpenClaw, but monitoring is still non-negotiable. The goal is resolving issues before the client notices downtime, which is the core of your 'unlimited support' promise and critical for client trust.
How do I onboard a new AI agent client in 48 hours?
Create an Orgo workspace and cloud VM for the client. Use your setup agent to deploy a Hermes agent inside the VM. Install Composio, Agent Mail, and an Obsidian vault loaded with their business context. Focus the first 48 hours on solving the one or two highest-pain tasks — do not overbuild. Ship the first working agent (e.g., inbox triage or follow-up drafting) within 48 hours. Set up Trello for ongoing delivery management and send a Loom walkthrough as the first delivery receipt.
How do I create content to attract AI agent clients?
Post content demonstrating agent capabilities specific to your target vertical before you have paying clients. Show real workflows: an agent triaging a law firm partner's inbox, an agent drafting insurance follow-ups, an agent managing a real estate lead pipeline. Use AI to assist with research and editing. The goal is that when a prospect books a call via your Calendly, they already know who you are and what you sell. Content is the highest-leverage warm-lead engine in this playbook.
How do I run a discovery call for AI agent services?
Lead with the Executive Abstraction Layer: ask about email overload, meeting burden, follow-up chaos, open loops, and scattered context. Every decision-maker relates to these pains instantly. Then transition to industry-specific questions — case management bottlenecks for law firms, lead follow-up gaps for real estate, claims processing friction for insurance. Frame every solution in terms of revenue generated or business outcomes, never 'hours saved.' Close by describing the unlimited digital employee offer at $5,000/month.
// Troubleshooting
What happens if my client's AI agent gateway crashes?
If your watchdog system is properly configured, the agent auto-detects the crash and attempts auto-restoration. Simultaneously, the agent emails you via Agent Mail alerting you to the failure. You resolve the issue remotely through your master Hermes agent's Orgo MCP access — often from Telegram while mobile. The client should never notice downtime. If watchdogs are not set up, the client discovers the failure first, which destroys trust in your unlimited support promise.
What if a client keeps requesting more than 1-2 changes per cycle?
Enforce the boundary through the Trello Kanban board structure. Requests live in Backlog; the client moves their priorities into To-Do. You pull 1-2 items into Doing per 48-hour cycle. If a client consistently requests more, have a direct conversation reframing expectations: the agent improves incrementally each week, not all at once. Remind them the unlimited offer means everything gets done — just in priority order. Without this boundary, fulfillment overwhelm will destroy your ability to serve other clients.
What if none of my initial test verticals show strong market pull?
If after 30-60 days no vertical shows clear inbound pull, referral patterns, or ease of delivery, re-evaluate your content angle and messaging before switching verticals entirely. Ensure you are leading with executive pain points, not technical AI features. If messaging is correct and pull is still weak, test a new set of 2-3 verticals. Do not stay in the Diverge phase beyond 90 days total — pick the best available option and converge, then refine from there.
Why should I avoid local Mac Minis for hosting client agents?
Local Mac Minis make remote management impossible at scale — you cannot SSH into a physical machine behind a home router as easily as spinning up a cloud VM. The blast radius of a hardware failure is also much higher: if the Mac Mini dies, every agent on it goes down simultaneously. Cloud VMs through Orgo are remotely manageable, sandboxed, individually deletable, and scalable. This is non-negotiable for a solo operator managing 8+ clients.
// Comparisons
How is this different from just reselling ChatGPT or Claude to businesses?
Reselling ChatGPT or Claude is a commodity play with zero defensibility — anyone can sign up for the same tools. The Orgo playbook sells a vertically specific digital employee that knows the client's business, people, and workflows through a structured Obsidian vault. You handle all infrastructure, monitoring, and iteration. The client never touches tokens or models. This specificity, combined with the unlimited offer and ongoing fulfillment, creates switching costs and premium pricing that generic AI reselling cannot achieve.
How does Hermes compare to OpenClaw as an agent harness?
Hermes is Nick Orgo's preferred agent harness because it is self-evolving, model-agnostic (you can switch underlying models without changing infrastructure), and has more stable gateways than OpenClaw. OpenClaw is an earlier agent harness now considered commoditized with less reliable gateway stability. While OpenClaw can work for $5K/month positioning, Hermes is recommended for the playbook because its stability reduces watchdog incidents and its model flexibility future-proofs your stack.
How does this compare to building a traditional AI SaaS product?
A traditional AI SaaS requires significant upfront engineering investment, product-market fit testing, and typically venture capital. The Orgo playbook is a productized service — you start generating $5,000/month revenue from client one with minimal upfront investment. You trade scalability ceiling for speed to revenue and capital efficiency. The playbook can also serve as a stepping stone: patterns discovered across clients can later inform a SaaS product if you choose to build one.
// Advanced
Can I charge more or less than $5,000/month?
The $5,000/month price point is calibrated to be affordable for SMBs while maintaining strong margins for a solo operator (since most clients need only 1-3 agents). Charging less undervalues the service and attracts clients who view AI as an expense rather than an investment. Charging more is possible once you have case studies and proven ROI in a specific vertical, but starting at $5K removes price objections while filtering for serious businesses. Do not itemize or discount — the unlimited framing is the offer.
How many clients can one person realistically manage with this playbook?
The playbook is designed for managing 8+ clients simultaneously using a master Hermes agent connected to all client Orgo workspaces via Orgo MCP. From a single Telegram chat, you can send tasks across all client VMs. At 8 clients and $5,000/month each, that is $40,000/month in recurring revenue as a solo operator. The 1-2 requests per 48-hour window delivery cadence, templated Obsidian vaults, and agents-build-agents model are what make this manageable without a team.
What does the master Hermes agent actually do day-to-day?
Your master Hermes agent acts as your operations manager across all client workspaces. Connected to every client Orgo workspace via Orgo MCP, it can fix gateways, upgrade Obsidian vaults, deploy new agent skills, check VM status, and run diagnostics — all triggered by natural-language commands from Telegram. You can send long-horizon tasks while walking, sleeping, or in meetings. A single query like 'how many Orgo VMs do I have across workspaces?' gives you a full operational overview. This is the fulfillment engine of the one-person business.
How do I handle security concerns from enterprise-adjacent clients?
Each client gets an isolated Orgo workspace with dedicated cloud VMs — no data crosses between clients. Composio handles authentication and tool calling with proper OAuth flows, eliminating credential sharing. Agent Mail uses named email addresses on the client's domain, keeping communication within their ecosystem. The Obsidian vault is per-client and sandboxed. Frame security in the sales conversation as: 'Your agent lives in its own secure environment, connected to your tools through enterprise-grade authentication, and only you and I have access.'
Should I build AI agents for my own business first before selling to clients?
Yes — using your own Hermes agent to manage your business operations serves as both proof of concept and practice. Set up your own Obsidian vault, connect Composio to your tools, and use Agent Mail for your business communications. This gives you firsthand experience with setup, maintenance, and troubleshooting that translates directly into smoother client onboarding. Your own agent also becomes your master operations manager as you scale — it is not optional, it is infrastructure.