Free Starter Kit
Free The foundation — 19 skills for memory, routing, and project setup
Stop prompting from scratch. 19 MIT-licensed Claude Code skills covering agent setup, cross-session memory, model routing, output mode, and session handover. The same foundation we use across every project.
- agent-setup-wizard — generates a CLAUDE.md that teaches Claude your codebase
- memory-ladder — 7-layer cross-session memory so the agent stops forgetting
- task-router — route each job to the right model before you start
- session-handover — clean state capture when you hit a context limit
QA Automation Pack
Coming soon Test loop, repair cycle, and coverage mapping — wired for Claude Code
A structured skill set for AI-assisted QA: red-green-refactor discipline, local model repair loops, ug-ug compressed test logs, and coverage gap detection. Pair it with the free kit for a complete dev + QA workflow.
- TDD vertical-slice discipline with AI repair loops
- Compressed ug-ug test logging (60-70% token savings)
- Coverage gap detection + auto-suggest test cases
- Escalation triggers for local → cloud model handoff
Designer Toolkit
Coming soon Figma-to-code handoff skills for AI-assisted front-end work
Skills for the design-to-build gap: component mapping, Figma annotation conventions, responsive token generation, and handoff docs that Claude can actually use as instructions.
- Figma component → Claude Code component mapping
- Design token extraction + responsive variants
- Handoff doc generator Claude can use as build instructions
- Visual regression prompt patterns
Done-For-You Setup
Service We wire your AI toolkit — you start shipping faster on day one
We spend 2 hours with your repo and your stack, then deliver a configured CLAUDE.md, project context files, a curated skill set, and a 30-minute walkthrough. You walk away with an AI setup that actually knows your project.
- Configured CLAUDE.md tailored to your stack and conventions
- Curated skill selection from the 400+ skill hub
- Project context files (WORKFLOW, LESSONS, OPEN-ITEMS)
- 30-minute walkthrough so your whole team knows how to use it
Maintenance Plans
Service Ongoing coverage for apps that need someone watching them
A monthly block of hours on your live app, so somebody is watching it between releases. The hours cover whatever the app actually needs that month — bug fixes and dependency and security patching first, but small changes and improvements fit inside the same block wherever there is room. Plans run $600 to $2,500 a month and every one of them is adjustable: tell us what your app needs and we will size the block to it.
Basic $600/mo 4 hrs/mo
- Bug fixes and dependency + security patching
- Uptime monitoring with alerting
- Backup verification — restores are actually tested, not just run
- Monthly email report: what changed, what broke, hours used
- Small changes and tweaks fit inside the hours where there is room
Standard $1,350/mo 10 hrs/mo
- Everything in Basic
- Small feature requests, not just fixes
- A monthly call to go over the app and what is next
Mad Scientist $2,500/mo 20 hrs/mo
- Everything in Standard
- A call every other week
- If we spot an easy win inside the hours and nothing else is queued, we build a prototype on our own time and bring it to you — you decide whether it ships, no obligation and no extra charge unless you want it built out further
- Month-to-month. Cancel anytime.
- Not a bug-only contract — unused hours go to small changes and improvements you actually want
- Every plan is adjustable; the three tiers are starting points, not a menu
Post-Production Maintenance
Service Already paying someone to keep your app alive? We will do it for less
For products somebody else built. We take over ongoing maintenance of your live app — patching, dependency upkeep, bug fixes and small changes — and most clients save 30 to 50% against what they pay now, depending on how much coverage they actually want. A plan that needs patching and monitoring costs far less than one absorbing steady feature work. The floor is $600/mo, and the offer works best if you are currently paying $1,500/mo or more. It starts with a call, then access to the repo and your ticket history. Once we have both, you get a real number within two business days.
- Typically 30–50% below your current spend, depending on the coverage you need
- Best fit if you pay $1,500/mo or more — the floor is $600/mo either way
- Our pricing assumes we can run our own automated tooling against your codebase; if your agreements restrict that, we quote it separately
- A short discovery call and repo access first, then a quote within two business days — revisited after 90 days once real ticket volume is clear
- Same scope as our maintenance plans: patching, fixes, small changes, monitoring
Vibe-Code Rescue
Service Built it fast with AI, now it needs a real developer — we finish it
You, or your AI tool, got the MVP most of the way there. We audit what is actually in the codebase, fix the parts AI coding tools reliably get wrong, and finish it to something you can ship or hand to a team without apologising for it. Scoped as a contract, because no two half-built apps are broken the same way — the audit comes first and the number follows it.
- Codebase audit — what is real, what is stubbed, and what is a security gap
- Fixes for what AI tools commonly skip: auth, payment webhooks, data validation
- The remaining stretch finished to a genuinely shippable state
- Roll straight into a maintenance plan once it ships, if you want it covered