Orchestrator
tech-lead
Takes the messy ask, breaks it into a plan, and hands each piece to the specialist that should own it. Your entry point for anything non-trivial.
Your AI engineering & marketing team in one command
89 specialist agents, 122 skills and 181 slash commands for Claude Code. Plan, build, ship and market a real product on your own.
Requires Claude Code · One-time payment · Lifetime updates
35 builders are already shipping with AgentsKit - all in the last 3 weeks. Only 15 seats left at $89.
↓- 89 agents engineer + marketing
- 122 skills
- 181 commands
- ~2 min to install
- No subscription pay once
Features
Load your and just prompt it.
Named specialists you delegate to — or let tech-lead pick for you.
Release gate
shipper
The last pair of eyes before anything goes out. Runs the checks, catches the half-finished work, and refuses the release when it isn't ready.
Backend
backend-architect
Designs the data model and the service boundaries before a line of it gets written, so you aren't unpicking the schema three sprints later.
Frontend
react-specialist
Components, state and rendering paths that survive a real app. Pairs with nextjs-developer and ui-ux-designer for full-surface work.
Firefight
debugger
Reproduces first, theorises second. Works the stack trace down to the actual line instead of rewriting code until the symptom moves.
Quality
code-reviewer
Reads the diff the way a senior would: correctness, then the shortcut you took at 1am, then the thing that will page you later.
Security
security-auditor
Threat-models the change and hunts the boring vulnerabilities — authz holes, injection, leaked secrets — with penetration-tester on call.
Data
postgres-pro
Indexes, query plans and migrations. Works alongside database-optimizer and sql-pro when the slow page turns out to be the database.
Infra
sre-engineer
Deploys, rollbacks, alerting and the incident path. terraform-specialist and kubernetes-specialist handle the substrate underneath.
Nine of 58 agents in the engineer kit. Cherry-pick one with agentskit add agent tech-lead, or install the lot in one command.
A team, not a tool
Nothing new to learn.
Picture briefing a tech lead, a backend architect, a security auditor and a copywriter on your project every single morning, from zero.
That is what prompting one generalist feels like. Every session starts cold, and the result depends on how well you remembered to ask.
AgentsKit installs the team into your project's .claude/ folder. Claude Code picks it up on its own, reads your CLAUDE.md once, and hands each job to the specialist that should own it. Watch it happen.
Your project
›_Your prompt
tech-leadplans
specialistsbuild
shippergates
feat: stripe checkout ✓Your repo
agentskit updateUpdatesfree, for life
Lives in .claude/
Native Claude Code format. No new app, no daemon, no account.
Reads your CLAUDE.md
Teach it your project once. Never re-explain it again.
Any stack
Next.js, Django, Rails, Go, Rust. Zero forced architecture.
Plain files you own
Markdown agents and skills. Updates never touch your custom files. How updates work
How it works
Two minutes, start to finish.
- 1
Buy
Pay once. Private-repo access lands the moment you check out.
- 2
Install
One
npxcommand. Pick engineer, marketing, or both. Nothing global. - 3
Prompt
Ask in plain English. tech-lead plans it, specialists build it, shipper gates it.
$
Two kits
A software team. A growth team.
Take one or both. They read the same codebase and the same CLAUDE.md, so the launch post knows what the feature actually does.
Engineer Kit
the software team58agents
61skills
159commands
- tech-lead plans and delegates, shipper gates every release
- backend-architect, react-specialist, typescript-pro, rust-pro, golang-pro
- postgres-pro, kubernetes-specialist, terraform-specialist, sre-engineer
- code-reviewer, test-automator, debugger, security-auditor
- Next.js, Tailwind, Drizzle, Docker, Stripe, MCP, TDD, Playwright skills
Marketing Kit
the growth team31agents
61skills
32commands
- growth-strategist finds the funnel constraint, brand-voice keeps copy on-brand
- seo-specialist, content-marketer, competitive-analyst, market-researcher
- /campaign-brief, /blog-post, /email-sequence, /landing-page, /launch-plan
- SEO audits, programmatic SEO, the full CRO set, pricing, attribution, PR
- New 19 skills added recently, incl. GEO/AEO and marketing-council
In practice
You type it once. It ships like a team built it.
You type
/create-feature add stripe checkout with webhooks, tests, and a launch post for the blog
You get
- tech-lead splits it into 4 tasks
- backend-architect schema, route, webhook handler
- test-automator 12 tests, all passing
- security-auditor 0 critical findings
- shipper ready to merge
- content-marketer launch post drafted in your voice
Compared
AgentsKit vs building it yourself.
Both end with a tuned Claude Code setup. One takes about two minutes. One takes a year of evenings.
| Job | AgentsKit | Building it yourself |
|---|---|---|
| A system prompt Claude keeps to | One CLAUDE.md, read by all 89 agents | 6+ hrs of trial and error |
| Agents that hand work to each other | tech-lead → specialist → shipper, wired | 2 days of wiring |
| Your stack and conventions | Taught once, matched on every task | Re-explained every session, 4+ hrs |
| Review and test workflow | /multi-agent-review, test-automator | 8+ hrs hand-rolling one |
| Release checklist | /deploy-checklist, shipper | 5+ hrs, rewritten per launch |
| CI, containers and infra | /setup-ci-cd-pipeline, docker-expert | 10+ hrs the hard way |
| Marketing and SEO frameworks | marketing-council, /seo-audit | 7+ hrs of hunting |
| Knowing the setup is any good | agentskit doctor, free updates | Never quite sure |
| Shaped exactly to your taste | Close, then edit any file | Fully, eventually |
| Time to a working team | ~2 minutes | 56+ hours |
Reviews
From people who bought it.
buying it last week was probably one of my best investments. genuinely love what he’s building here 🔥
abhiFounder of craftpad · @letcontactabhi
See it on X ↗
Pricing
No subscription.
Pay once, get private-repo access the moment you check out, and every future update after that. No seats, no renewal, nothing to cancel. Works with any Claude Code plan: Pro, Max, Team or API.
- Both kits: 89 agents, 122 skills, 181 commands
- Both CLAUDE.md templates + the agentskit CLI
- Private repo, lifetime updates
- 14-day money-back guarantee, no questions asked
Bought apart, the two kits are $98. Just need one team? See the single kits below.
Complete Bundle
$89
15 launch seats left
once, lifetime updates
Get the BundleGoes to $149 after 15 more sales
Secure checkout via Polar · instant private-repo access · 14-day money-back guarantee
Engineer Kit
the software team · 58 agents, 61 skills, 159 commands
Marketing Kit
the growth team · 31 agents, 61 skills, 32 commands
15 seats left at $89 — after that the bundle is $149.
Very few available · only 50 seats ever at $89
35 of 50 claimed at $89First 50 customers only · then $149
$89
first 50 · 15 left
You are here$149+$60
after 50 customers
Launch pricing. 15 seats remain at $89, the next buyer after that pays $149.
FAQ
Questions people ask.
Do I need Claude Code? Does it work with Cursor or ChatGPT?+
Yes. AgentsKit is built specifically for Claude Code, Anthropic's AI coding tool for the terminal, desktop app, and IDE. The kits are agents, skills, and slash commands that live in your project's
.claude/ folder, which is a Claude Code format. It does not run inside Cursor, Copilot, or the ChatGPT app. Any Claude Code plan works: Pro, Max, Team, or API billing.What exactly do I get?+
Access to a private GitHub repo containing the kit(s) you bought: a
.claude/ directory of agents, skills, and slash commands, CLAUDE.md templates, the agentskit installer CLI, and full docs. You drop it into any project with a one-line npx github:getagentskit/… command that pulls straight from your private repo. The exact command is in your repo's README.Do I need to know how to code?+
You need Claude Code and a project to work in. The agents do the heavy lifting. You direct them in plain English. Installation is a single command.
Does it lock me into a framework?+
No. Agents adapt to your stack (Next.js, Django, Rails, Go, Rust, anything) by reading your CLAUDE.md. Zero forced architecture.
How do updates work?+
Buy once, get every future update. Run
agentskit update (or git pull) to refresh. Your custom files and CLAUDE.md are never touched.Is this affiliated with Anthropic?+
No. AgentsKit is an independent, unofficial product and is not affiliated with, endorsed by, or sponsored by Anthropic. "Claude" and "Claude Code" are trademarks of Anthropic.
Is there a money-back guarantee?+
Yes. 14 days, full refund, no questions asked.
Put AgentsKit on a real feature. If it doesn't save you hours on that first ship, email us within 14 days and we'll send the money back.
We can offer that because the kit isn't theory. Every agent, skill, and command in it earned its place in real production work before it shipped to you.
Still stuck? Email us. Want to see it first? Watch the demo.