/ˈeɪdʒənt ˌhɜːr.dər ˈbuːtˌkæmp/ (noun) — the productized 6-week prep program for engineers who want to stop running one Claude Code session at a time and start running a fleet on real work.
Most engineers can get to two or three parallel Claude Code sessions on their own. The jump from three to a working fleet of eight or more is where the practice actually begins — and where most people stall.
The Bootcamp is the structured path through that wall. Six weeks. Self-paced. Real assignments on your real codebase. By the end, the practice is just how you work.
Reserve your seat → $995 · 6 weeks self-paced · Outcome-guaranteed
We asked engineers running Claude Code what was actually getting in their way. Three answers came back:
"I've gotten two or three sessions working, but past that it falls apart."
"Every YouTube video shows a toy repo. I need this on my real codebase, with real merge conflicts and real CLAUDE.md."
"Skills, hooks, sub-agents, plugins, MCP, mission files — what order do I actually learn this in?"
That's not a you problem. That's a missing-curriculum problem.
There's no book. No conference talk. No course. The handful of engineers running 10+ parallel agents today figured it out by running into walls and writing their own tooling. The Bootcamp is the path they didn't have — assembled, sequenced, and tested against real work.
You're not behind. You've been trying to learn a practice that didn't have a curriculum yet.
Each week has a single concrete deliverable. By the end of the week, you've shipped that deliverable in your real repo. No toy projects. No throwaway demos.
cctabs installed in your terminal. First parallel session running. Your repo's CLAUDE.md configured to brief any new agent in under thirty seconds. By end of week you're orchestrating from one window instead of switching between fifteen.
One agent, one task, one repo, feedback loops working. The Level-1 baseline most engineers can self-teach but rarely teach themselves deliberately. You'll brief an agent so well it ships a real feature without you intervening.
Three to five agents on separate branches. Coordination patterns that don't collapse under merge conflicts. This is the wall most engineers hit — and the week that actually moves you past it.
Build and install your first three personal Claude Code skills. Skills are how you teach Claude your workflows — not generic advice. You'll author skills tied to the daily work you actually do.
Eight or more parallel sessions running on real work, briefed via mission files, with cross-tab coordination. This is the Level-3 practice. By end of week the fleet is real, not theoretical.
The practice integrated into your daily tools, your sprint cadence, your task system. The "this is just how I work now" stage. By end of week you're not running a fleet for the Bootcamp anymore — you're running a fleet because that's the job.
There's a particular failure mode for engineers learning Claude Code today. You watch tutorials. You read other people's repos. You skim documentation. You take notes on someone else's setup and intend to try it next sprint. It feels like learning. Most of the time it isn't.
The science of how skills actually become durable is unforgiving:
Retrieval beats re-reading. Watching someone else run a fleet doesn't build a fleet. Running one on your codebase does. Every week of the Bootcamp ends with you shipping a deliverable on your real repo — not consuming more material.
Spaced beats crammed. A weekend bootcamp is the worst possible architecture for skill durability. Six weeks of one-pattern-per-week is roughly the right size for the practice to embed in your daily work, not stay in a Notion doc you never reopen.
Generation beats showing. Each week's curriculum has you attempting the work first, then walking through how I'd approach it. Trying — and getting stuck — before being shown the answer is what makes the lesson stick. Skipping straight to the answer feels efficient. It isn't.
Real difficulty beats safe practice. Toy repos that always work don't build engineers. Your real codebase, with its real merge conflicts and its real CLAUDE.md weirdness and its real production constraints, is what pushes the practice through the plateau.
This is why six weeks, not a weekend. Why your real repo, not a sandbox. Why each week has one concrete deliverable, not twenty checkboxes. The architecture is doing work the curriculum content alone can't.
The 6-week curriculum is the core. But every working asset I've built running this practice for myself and for the engineering teams I coach is also in the box.
Bonus #1
Eight documented end-to-end case studies of AI doing actual work, with the prompts, outputs, and lessons. Medical report translations across languages, navigating a Finnish healthcare portal in a language you don't speak, shareholder-loan forensics in Procountor, timesheet automation against an enterprise portal, a customer-service negotiation thread that actually got the refund, the biography I wrote for my dad's 75th. Not "10 ChatGPT prompts." Real life with receipts.
Bonus #2
Production-ready Claude Code skills, packaged for install. Code review. Security review. Mission planning. Skill authoring. The herd CLI plugin. The skills running in my own setup, ready to drop into yours.
Bonus #3
Command patterns and recipes for the daily fleet workflow. The unwritten parts. How to fork a stuck agent. How to consolidate a five-agent investigation across mission files. How to keep mission files clean. How to recover when an agent goes off-course without losing its work.
Bonus #4
The briefing format that turns "go fix this" into agent-executable work. Includes the actual mission file I used to ship the recent agentherder.com rebrand across four parallel agents — annotated, with the parts that mattered called out.
Bonus #5
What AI-native engineering adoption actually looks like inside cybersecurity firms. Drawn from coaching engagements with F-Secure and DNV Cyber. What worked. What didn't. What you'd want to know before pitching it inside your org — and the responses to expect from security, procurement, and senior management.
Included
Three months of the Club included with the Bootcamp. Monthly drops of new skills, mission file patterns, and case studies. After 90 days it's optional and continues at $37/mo (or $97/mo Engineering tier).
Plus the AgentHerder Certification track. The Bootcamp prepares you for the certification exam — the timed, transcript-graded credential awarded by a Principal AgentHerder. (Cert exam fee separate; see Pricing.)
Six weeks. Eight parallel sessions on real work. Or full refund.
Show up. Do the assignments on your real repo. If at the end of the program you can't run eight or more parallel Claude Code sessions on real work — refund, no questions.
This is outcome-bound, not satisfaction-bound. Not "if you didn't enjoy the videos." If the practice doesn't take, your money is back. The bar is the same one I hold myself to running 12–15 sessions daily.
Provisional pricing for the launch cohort. May rise once the first cohort completes.
Individual
$995 one-time
Team
$3,995 5 seats
After the included 90 days, AgentHerder Club continues as the recurring layer:
Club is optional. The Bootcamp doesn't gate on it.
The certification exam is sold separately and graded by a Principal AgentHerder. See the certification section for details.
No. Solo indies, individual contributors at companies, and engineering managers all complete the program. The fleet you'll build is yours, on your repo. Team adoption is a different conversation — happy to talk about that separately, but it's not what the Bootcamp is for.
Yes — arguably better than for someone in a corporate environment. No procurement review on every new tool, no security team blocking outbound traffic, no compliance call on whether agents can touch the codebase. The Bootcamp's curriculum is designed around an engineer who can ship into their own repo without asking permission.
Week 1 assumes you've used Claude Code at least once and have a working terminal setup. If you've already run two or three sessions, Week 1 will feel light and the real work starts in Week 2. Total beginners can still complete the program but should plan for extra time in Weeks 1 and 2.
cctabs currently runs in Wave Terminal on macOS and Linux. Windows works under WSL2. Native Windows is on the roadmap. If you're on Windows today and don't have WSL, the Bootcamp will be harder — possible, but harder.
Mostly yes, with caveats. cctabs is npm install -g. Claude Code needs either a Claude.ai subscription or an Anthropic API key. Some enterprise environments block outbound traffic to Anthropic's API or restrict global npm installs — check with your IT team before enrolling. If you can't get Claude Code running on your work laptop, you can do the Bootcamp on a personal machine.
Both. ICs build a personal fleet they own. Engineering managers learn what the practice looks like from the inside, which is the prerequisite to evaluating it (and to evaluating engineers who claim they're already doing it).
Claude Code (Claude.ai subscription or Anthropic API key — your choice). A code editor. A terminal that runs cctabs (Wave Terminal currently). A GitHub or similar git host. Optional but useful: Linear or your team's task tracker for mission files. That's it. No other paid software is required.
Fully async. No live calls. No scheduled cohorts. No mandatory Zooms. You can buy any time and start any time. Cohorts (people who start in the same month) get a private channel together — community signal without facilitation overhead.
Take Week 6 again. Or take the whole program again — your access doesn't expire. Or trigger the guarantee. The point is the outcome, not the calendar.
That's roughly how long the Solo → Parallel → Fleet progression takes when you do the assignments deliberately on a real repo. Faster is possible if you're already at Level 2 and have time to push hard. Slower is fine — the program doesn't lock you out.
The cert is sold separately. The Bootcamp is the prep program — it gets you cert-ready. The cert is the credential. You can take the cert without the Bootcamp; you can take the Bootcamp without the cert. Most people will do both.
The Augmented Mind Bootcamp is the on-ramp — for people who use ChatGPT or Claude a few times a week and want AI as a steady presence in their everyday life. AgentHerder is the destination — for engineers who want to scale into a fleet of parallel coding agents on a real repo. If you're not yet running parallel sessions, start with Augmented Mind. If you're an engineer plateauing at 2–3 sessions, you're in the right place.
Fredrik Wollsén. See "About" below.
Built by Fredrik Wollsén — Agentic Software/Growth Engineer at Augmented Mind. Twenty-two years as a professional software engineer; the last decade making AI genuinely useful, personally and professionally.
Currently running 12–15 parallel Claude Code sessions every day through cctabs — the open-source CLI he built for the job. Coaches engineering teams at cybersecurity firms (F-Secure, DNV Cyber) on AI-native engineering practices. Author of The AI-Native Engineering Playbook: Crawl, Walk, Run, Fly (1,100+ reads). Founder of Remember This and My Transcriber. Previously Mozilla.
On stage at: GOSIM AI Vision Forum Paris, Panel 1 — Agentic AI Systems & Human-AI Symbiosis (May 4, 2026).
Everything in this Bootcamp is something the curriculum author does every day. Not theorised. Not from a deck. From actually running the practice.
Six weeks. One curriculum. Your real repo. Outcome-guaranteed.
The Bootcamp opens for its first cohort after Paris (early May 2026). Get on the early-access list — we'll email when enrollment opens, and the first 50 enrolments get launch pricing locked.
No spam. One email per milestone. Same waitlist as the certification — you'll be notified about both.
Not running parallel agents yet? Start with the Augmented Mind Bootcamp — the on-ramp for everyday AI use, before you scale into a fleet.