Skip to content

theluckystrike/theluckystrike

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

72 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

michael lip — fractional ai engineer · 10,000+ users in production · 20 extensions shipped · 47,886 commits · 140 prs upstream · 99.9% uptime · 4.6/5 avg rating · 1 slot open q3 · lipmichal@gmail.com

profile view counter

ship.ts — const cycle = await pipeline({ spec: writeSpec(client.brief), tests: generateEvals(spec, { framework: 'vitest' }), agent: deployAgent(spec, { model: 'claude', mcp: true }), review: pairSession(client.engineer), ship: deploy({ target: 'cloudflare', rollback: '<60s' }) }); assert(cycle.evals.pass); assert(cycle.uptime >= 0.999); // handoff.md written. your team maintains.

the pitch in one line. i embed with seed/series-a saas teams as their first ai engineer. ship agents, mcp servers, eval pipelines in 14-day cycles, then hand off a repo your team can actually maintain. typed everywhere. tested everywhere. claude code drives the keyboard. i drive claude.


§ 01 · proof of work

proof of work — tab-suspender-pro 200+ installs/day 4.6 rating BSL-1.1 · belikenative mcp grammar server 70 rules 10k+ users 4.6 rating MIT · chrome-extension-toolkit typescript-first mv3 scaffolding 12+ shipped extensions · content engine 4,200+ articles daily cadence 1,700+ views/mo

tab-suspender source · belikenative action · extension-toolkit · claudecodeguides.com


§ 02 · the retainer — what most clients want

fractional ai lead embedded retainer · $10-12k/mo · 6-month minimum · 30-day pilot with full refund · 1 slot open Q3 2026

why a retainer beats a project for ai work — ai roadmaps shift weekly, evals and prompts need maintenance, your team learns by pairing, renewing is cheaper than re-hiring — the deal: $10-12k/mo 6-month min 30-day refund, capacity 3 retainer clients max, currently 2 active 1 slot open Q3 2026, you own the repo the evals the runbooks


§ 03 · also available · one-shot

agentic mvp one-shot — 21-28 days — from $12k — 28-day refund chrome extension build — 14-21 days — from $8k — store-approval guarantee ai architecture review — 1 week — $6k flat — credits toward retainer
the architecture review is the cheapest way to test fit.
$6k. one week. written report + diagram. credits to retainer if you sign.

§ 04 · who this is for
  [ok] seed -> series-b saas. 5-80 engineers. ai is on the roadmap or already shipped-and-stuck.
  [ok] founders who'd rather rent senior judgment than gamble on a $250k full-time hire.
  [ok] teams shipping to real users. production-grade evals matter more than demos.
  [no] pre-product startups looking for a co-founder. i build, i don't bet equity.
  [no] enterprises needing 6-month procurement cycles. i'm one person, paid in stripe.
  [no] "build me a chatgpt clone" briefs. i need a real problem.
§ 05 · how the work runs · agentic engineering by default
  step 01   spec.md           -> written. signed. shared repo. success metrics named.
  step 02   failing.test.ts   -> vitest + playwright + eval harness before code.
  step 03   typed.code        -> ts everywhere. zod at every boundary. no any, no maybe.
  step 04   ci.green          -> no merge, no ship, no exception.
  step 05   deploy.sh         -> cloudflare. <60s rollback path. you have the keys.
  step 06   pair.day          -> weekly. one of your engineers pairs with me on review.
  step 07   handoff.md        -> runbooks. ADRs. eval failure playbooks. you maintain.
§ 06 · how i work — communication and commitment
this is my full-time job. my only income. i treat it accordingly.

communication:
  -> written-first. always. no exceptions.
  -> comprehensive reports after every sprint — you'll see exactly
     what was built, why, what changed, and what's next.
  -> async by default. i'm gmt+7, you're probably not.
  -> strongly prefer long-form written briefs over calls.
     if a call is needed, i'll ask — but it rarely is.
  -> every decision documented. every trade-off surfaced in writing.

commitment:
  -> this is not a side hustle. zero other clients beyond capacity.
  -> your project gets deep focus, not scattered attention.
  -> i ship daily. you see progress daily.
  -> if something is blocked, you know within hours — not days.

what you get that's different:
  -> reports that read like engineering docs, not status updates.
  -> zero meetings that could have been a message.
  -> a written paper trail you can search, reference, and hand off.
§ 07 · receipts
$ git log --oneline --author=theluckystrike --since="12 months ago" | head

  a3f8c12   feat(tab-suspender): mv3 service-worker rewrite · ship v2.1
  91d2e44   merge:  axios#6291 · keepAlive defaults restored
  b7e019a   merge:  google-chrome/chrome.dev#142 · ext-page typings
  4f1c8b3   feat(belikenative): mcp server · 70 rules · 18 l1-aware
  ce6a201   merge:  nicehash/nh-monorepo#88 · worker memory leak
  8dbf17f   feat(zovo-toolkit): wxt + plasmo dual scaffold
  2a4e9c1   merge:  microlinkhq/preview#56 · cls regression patch
  efd03b8   chore:  gh-actions · grammar-check · 60->70 rules
  19b6f72   merge:  scriptscat/scriptcat#214 · cookie-store typings
  c7b1f0d   feat:  mv2->mv3 migration #12 · zero rank loss
  ...
  total 12mo:  47,886 commits · 140 prs merged upstream · 0 regressions reverted

why open-source receipts matter at this stage. i'm transparent: most of my deployed work is mine — 20 extensions, 10k+ users, 4.6 avg. i'm building a paid client portfolio now, and the trade is honest: hire me before the testimonial wall fills up, get senior shipping speed at a junior-consultant price. month-1 pilot is refundable for a reason.


§ 08 · stack

tech stack — agentic: claude code, mcp, gemini cli, minimax, langgraph, vercel ai sdk — frontend: typescript, react, next, vite, tailwind, shadcn — backend: node, python, django, supabase, postgres, trpc — extensions: mv3, service workers, dnr, wxt, plasmo — infra: cloudflare workers, aws, docker, github actions — testing: vitest, playwright, msw, zod, eval harnesses


GitHub Stats

GitHub streak stats for theluckystrike — current streak, longest streak, and total contributions

47,886 contributions in the last year — theluckystrike contribution graph

47,886 contributions in the last year — organizations: @microlinkhq @BeLikeNative @scriptscat and more — contributed to chrome-tips, claude-skills-guide, ai-tools-compared and 752 other repositories


Latest Posts


§ 09 · where to find me

where to find me — lipmichal@gmail.com reply within 1 business day · zovo.one 20 chrome extensions 10k+ users · belikenative.com grammar tools for devs MIT open source

connect — zovo.one · @_alphashark_ on X · Upwork Top Rated Plus · LinkedIn


§ 10 · apply

apply for a Q3 retainer slot — lipmichal@gmail.com — reply within one business day

reply within 1 business day — yes, no, or scoped proposal — no discovery call until we both agree it's worth one — if the fit is wrong, i'll refer you to someone better — gmt+7 da nang vn zovo.one @_alphashark_

-> Apply for a Q3 retainer slot via email


$ ls ./background/

cto-turned-solo-dev. 10+ years building software, leading infra, shipping product. founder of zovo and belikenative. three content sites, 4,200+ articles. da nang since 2022.

  contributions.last_year   47,886
  current.streak            64 days
  followers                 16
  organizations             trustwrx · belikenative

built solo · shipped tested · lipmichal@gmail.com

About

Config files for my GitHub profile.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors