Store

This is the field guide I wish existed after my first production MCP deployment. The failure modes are not obvious from the docs: stdout pollution that silently kills your tool responses, concurrency races when two agents hit the same server in the same tick, and prod gates that return success codes while actually doing nothing. Every chapter is grounded in real shipping lanes — no theory-only sections, no padded word count.

The Playbook covers four phases: initial wiring and transport selection, the first week of real load (where stdout and concurrency problems surface), the second-week hardening pass (structured logging, idempotency patterns, health probes your CI can actually test), and the month-two plateau where teams discover their prod gates were always lying to them. Each phase ends with a copy-paste checklist you can drop directly into a CI workflow or a pre-launch review doc.

Who it's for: operators who have already shipped at least one production MCP and hit the wall — not developers evaluating whether MCPs are worth pursuing. If you are still in the evaluation phase, the Power Pack is the better starting point. If you are already past the wall and need implementation time, look at the time-block install instead.

A template repository plus setup guide for senior solo operators running Claude Code in production. The repo ships with a working.claude/scaffold: skills wired to real workflows, hook recipes tuned to the failure modes solo operators hit (dangling worktrees, leaked subagent writes, batch_commit staging more than you expected), and multi-agent dispatch patterns with concrete file-ownership matrices that prevent merge conflicts without requiring a manual review queue.

The guide that ships alongside the repo is not a README. It explains the reasoning behind each configuration choice, covers how to extend the dispatch patterns for your own domain, and walks through a worked example of dispatching parallel subagents against the same codebase without ending up with three agents rewriting the same file in different directions. The file-ownership matrix pattern alone has saved more rebases than I care to count.

Who this is NOT for: toy demos, beginner explorations, or anyone whose current production system is a single conversation with no tooling around it. The Power Pack is not a tutorial; it assumes you already understand how Claude Code works and want to run it at the ceiling of what a solo operator can sustain.

Pre-paid, scheduled, scoped to a single shipping outcome. You pick the block size that fits the scope, pay in full at checkout, and we book a video call. The call runs the full duration: MCP wiring, agent workflow configuration, buyer ops setup, or whatever specific outcome we agreed on when you booked. No retainers. No ongoing relationship unless you book another block.

The 2-hour block suits a single MCP integration or a focused workflow audit — enough time to wire one server, validate it against a real agent loop, and leave you with working code plus a summary doc. The 4-hour block handles multi-server setups or an end-to-end agent pipeline from cold start to CI-gated production. The 8-hour block is for operators who want to hand off an entire workflow layer — all MCP wiring, all agent dispatch configuration, all tooling hooks — and come out the other end with a running system they can maintain themselves.

Cancellations made 24 hours or more before the scheduled slot get a full refund. Same-day cancellations do not. Rescheduling is always free with 24 hours' notice. Async follow-up questions after the call are answered within one business day in BC, Canada (Pacific Time).

The Playbook, the Power Pack, and a scoped install window — packaged together at a lower combined price than buying each product separately. The bundle is designed for operators who already know they need all three: the conceptual grounding from the Playbook, the working scaffold from the Power Pack, and dedicated time to wire it all into their specific environment with someone who has shipped it before.

The install window in the bundle is scoped to getting your Power Pack scaffold running in your actual codebase — not a generic demo setup. You come with your repo, your MCP targets, and your existing agent configuration (or the absence of one), and we leave with a working system. The Playbook serves as the reference layer for decisions we make during that session; the Power Pack template becomes your starting point rather than a blank canvas.

If you are not sure whether you need the bundle or just the Playbook plus a time-block, the honest answer is: buy the Playbook first. Read it. If you hit chapter three and immediately recognize three problems you already have in prod, the bundle price will make sense. If chapter three describes problems you have not hit yet, the Playbook alone is enough for now.

Flat-fee wiring when your network needs a buyer channel. This is not a community management service — it is a one-time infrastructure setup: server structure, role topology, channel layout, and bot hook wiring. You come out with a Discord server configured for a buying audience, not a general community, and a runbook your team can operate without calling me back.

The bot hooks cover purchase events (Polar webhook integration), automatic role grants based on what a buyer purchased, gated channels that only unlock for the right role, and a basic moderation layer that keeps noise out of buyer-only spaces. The deploy script is self-contained and documented; the env-var checklist tells you exactly what needs to be set in your bot host before it goes live. Pricing scales with server complexity — a single-product one-tier setup lands near the floor; a multi-product multi-role buyer server with custom bot behavior lands near the ceiling.

This product is side-band only: it is available when I have capacity and the setup scope fits a flat-fee engagement. If your Discord needs are ongoing (content scheduling, community management, moderation ops), this is not the right fit. The handoff doc that ships at the end of the engagement is written for your community manager, not for me — the goal is that you never need to call me again for Discord maintenance.