Income: $1,000 USD/month recurring, at 20 hours/month
Sell fixed-price micro-builds to cold buyers on Upwork (Contra in parallel at $0), and stack keep-it-running retainers. The offer: "I direct Claude Code to ship your small web app, internal tool, or automation in 7 days" — $300–800 per build, $100–250/month retainer. Two builds a month. The arithmetic that forces this: $1,000 ÷ 20 h = $50 per effective hour, which no hourly job available to a pre-Step-1 student in Cairo pays, and which fixed-price builds reach because Claude Code does the typing.
Hard constraint: first money received before Sep 20 (his rule, 2026-08-22). Target Sep 5; two tripwires (Sep 7, Sep 14) below force an offer change and then a backstop gig so the date holds. Month 1 at 5 h/week is $300–1,000 one-off, not recurring (estimate, 2026-08-22). $1,000/month recurring is a month-2–3 outcome from 4–6 retainers plus repeat clients.
What a week looks like
Week 1 is front-loaded: 20 proposals in 7 days (≈4 h), because time-to-first-dollar on a cold platform scales with volume sent, not weeks elapsed. Then the steady state:
Sunday 21:00, 2 h. Open Upwork, filter "posted in last 24 h" for small web app / internal tool / automation jobs. Send 8 proposals, ≤150 words each, every one opening with a 60-second Loom of a live app you already built (sf-design.pages.dev, sf-course-c4h.pages.dev, signal-hud.pages.dev) and closing with a fixed price and a delivery date.
One evening mid-week, 3 h. The job you won: write a one-page spec, direct Claude Code to build it, run the verification yourself, catch what it got wrong, deploy to Cloudflare Pages, record a 60-second walkthrough, invoice. That session is a course rep — the same spec → direct → verify → ship loop the 120 h course trains — with a stranger's deadline attached.
Month 2 onward. Each shipped build is offered a retainer: hosting, fixes, one change a month, ≤1 h of your time. Four to six of those is the $1k.
Objective and done criteria
Objective. $1,000 USD/month recurring from cold buyers within 3 months, at ≤20 h/month, spending ≤$50 total, without touching the warm market (NGU students, doctors, family — reserved for tutoring and the app), with every paid hour doubling as a build-literacy rep.
| Done when | Evidence | By |
|---|---|---|
| First proposal sent | Upwork "My proposals" page shows one, timestamped | the day this is approved |
| First invoice paid | Upwork/Payoneer balance | target Sep 5 · hard: before Sep 20 |
| 2 builds delivered | 2 live URLs, 2 paid invoices | Sep 22 |
| ≥4 active retainers ⇒ ≥$1,000/month | retainer invoices paid in the same month | Nov 1 |
Non-goals. No warm-market outreach. No dashboard, CRM, or tracker — the Upwork inbox is the system. No course purchases. No salaried or hourly remote job. No ads. No selling the NGU-derived Anki decks (rights undefined, and they are StudyFix's own product).
Constraints, as stated
| Constraint | Value | Source |
|---|---|---|
| Target | $1k USD/month recurring | his answer, 2026-08-22 |
| Hours | ≤20 h/month total (~5 h/week) | his answer, 2026-08-22 |
| Capital | $1,500 total, earmarked for Step 1 / residency; "can't be in vain"; ≤$200 only for something near-certain | his answer, 2026-08-22 |
| Payment rails | Payoneer/PayPal working · USD bank account · relative abroad — not a constraint | his answer, 2026-08-22 |
| Market | Cold only. Warm market reserved for tutoring + the app "because it's my reputation" | his answer, 2026-08-22 |
| Skill goal | Earn while building the skills for "an app everyone uses to study"; open to a remote job | his answer, 2026-08-22 |
| Calendar | 120 h course D1 Aug 16 → D10 Aug 25 at 12 h/day, Module M after; The Eye Aug 17–24; Year 4 ~Sep 20 with a 2.5–3 h daily commute | victor\MEMORY.md L78–87 · _briefings\agent\MEMORY.md L13–21 |
| Failure mode to design against | Builds systems, ships 0–1 units. Starting is the bottleneck | _briefings\agent\MEMORY.md L49–62 |
What the disk already proves
Assets (sellable proof)
- ~20 deployed single-file study apps,
C:\Users\BYJar\Desktop\Sites\(Oct 2025 → Aug 2026), plus client-shaped one-offs (mumCPE,mamaplan,kevz,closer,routinemaker) - Three live apps built by directing Claude Code: the 120 h course, The Eye (28 step screens, 20 widgets), the Signal HUD; a PWA at gymmoi.thestudyfix.com
- A Telegram relay + responder + watchers stack (
_briefings\agent\) — a working automation, demonstrable - Prior paid work: marketing-agency appointment setter, real-estate agent, tutor; wrote and ran a car sale on Dubizzle (his words, 2026-08-22)
- Reads and watches Hormozi — offer design is not a cold start
Gaps (design around, don't hide)
- Zero record of ever charging anyone; no client, rate, or invoice on disk
- Waitlist = 1 (
plans\handoffs\HANDOFF_standalone-05f4619a_jarvis-week-plan_2026-08-17.mdL358); no audience on any platform - Observership outreach infrastructure exists with 0 delivered sends — proof that "build the console" is the trap
- Lit-review sprint at 0/20 h (
research\STATUS.md) — not a skill to sell this month
Why this path and not the others
to StudyFix
| Path | $/h at 20 h/mo | Month-1 $ | Recurring mechanism | Transfer /10 | Verdict |
|---|---|---|---|---|---|
| Fixed-price builds + retainers | $40–100 | $300–1,000 | retainers stack; repeat clients | 9 | chosen |
| Real-estate cold caller / ISA / VA | $4–8 | $80–160 | hourly only; needs ~165 h/mo for $1k, 4 pm–midnight Cairo | 2 | rejected — a skill he already has |
| Appointment setting, commission | volatile | $0–500 | requires 10–20 h/week of DM volume | 6 | rejected on hours |
| Medical writing / lit reviews | $25–60 | $0–400 | per paper; Gulf doctors pay USD | 3 | rejected — 0/20 h, slow first dollar |
| Online tutoring | $15–30 | $200–500 | hourly | 4 | rejected — below the bar, and his reserved market |
| Sell the Anki decks / indices | — | — | — | — | blocked — warm market, lecturer-content rights undefined |
Rates are my 2026-08-22 estimates from the freelance market as I know it; the Market agent replaces every one with a sourced 2026 figure or marks it unverified.
The engine
Bidding is pinned to one slot so it never competes with the course or Year 4 for a decision. Build hours are course reps.
$300–1,000
~$800–1,600
retainers ≥ $1,000
The offer, priced
| Item | Price | Scope |
|---|---|---|
| 48-hour wedge (the first-dollar product; also listed on Contra and Fiverr at $0) | $80–150 | one interactive page from the client's PDF or notes — quiz, flashcards, index, tracker — delivered in 48 h; his existing pipeline, ~1.5 h of his time. A low price buys a fast yes and the first review, which unlocks the builds |
| Paid trial build (first client only) | $150 | one screen or one automation, 3 days, one revision round — buys the first review |
| Standard build | $300–800 | single-file web app, internal tool, or automation; 7 days; one-page spec signed before work; one revision round |
| Keep-it-running retainer | $100–250/mo | hosting, fixes, one change per month, ≤1 h of his time |
Proposal template (the real shape — ≤150 words)
[Client's first name] — I build exactly this kind of tool. 60-second walkthrough of one I shipped
last week: [Loom link to sf-design.pages.dev walkthrough]
What you'd get: [restate their ask in one line, in their words]
How: one-page spec you approve on day 1 → working version on day 4 → your fixes → live on day 7.
Price: $[X] fixed. Date: [7 days from today]. Hosting included for month 1; $[Y]/mo after if you want me to keep it running.
One question so the spec is right on day 1: [the one thing their post leaves ambiguous]
— BetraDecisions
Hourly exposes that Claude Code does the typing and caps him at market hourly rates. Fixed price pays for the outcome and makes speed the margin. Scope is locked by a one-page spec the client approves on day 1; one revision round is included, more is a new build.
Upwork is the largest cold market with rails that pay to Payoneer in Egypt. Contra runs in parallel at $0 commission. No DMs to NGU, doctors, or family — that market waits for the app, on his instruction.
Connects are the one spend with direct attribution (a proposal sent). The Sep 20 date needs ~40 proposals by then at roughly $1.50–2.50 each in connects (unverified — Market agent confirms live pricing and new-account free connects); $50 would stop the volume at about week 2. Anything past $100 goes to Contra at $0 instead. No courses: the 120 h course already covers the build skill, and metalearning doctrine forbids buying the machine before doing reps. No ads: there is no offer to put money behind yet.
Four built-not-shipped trackers are on file. This plan adds zero: the Upwork inbox is the pipeline, one line in C:\Users\BYJar\Desktop\workspace\_briefings\REMINDERS.md holds the Sunday bid slot, and progress is counted only as proposals sent, jobs won, invoices paid. Profile polish is capped at 1 h and the first proposal goes out before the profile is "done".
Each build is the course's spec → direct → verify → ship loop with an external deadline, which is the forcing function the course cannot supply. Victor coaches the rep, never builds it (victor\MEMORY.md L8–11). How Victor logs them is Victor's call; the course content does not change.
Skills to learn now — provisional scoring
Provisional My judgment as of 2026-08-22, written so he can react now. The Skills agent (below) replaces this table with a sourced version that reads the actual course syllabus
(C:\Users\BYJar\Desktop\workspace\victor\course\index.html, D1 "JavaScript, all of it" → D10 "Ship it", Module M "The Market") and the learner doctrine in victor\FEEDBACK.md.
| Skill | End goal /10 | $ this month /10 | Course covers | Learned on this path | When | Already has |
|---|---|---|---|---|---|---|
| Directing Claude Code + catching false greens | 10 | 9 | D4–D10 (bluff drills) | yes — every build | now | Max plan, the course, 20 shipped sites |
| Scoping: vague ask → one-page spec | 9 | 9 | partial (D9–D10) | yes — every client | now | — |
| Offer design + copywriting | 8 | 9 | Module M, pricing only | yes — every proposal | now | Hormozi material, the Dubizzle ad |
| Pricing / packaging | 7 | 8 | Module M partial | yes | now | — |
| Customer discovery (what buyers actually ask for) | 8 | 5 | no | yes — every client call | now | — |
| TS / Next.js / Supabase literacy | 8 | 4 | D1–D5 | partial | now, via the course | the course |
| Prompt / pipeline / eval design | 9 | 3 | D6–D8 | no | now, via the course | the course, thestudyfix\prompt-lab\ |
| UX eye | 7 | 6 | The Eye | yes | running | sf-design.pages.dev |
| Distribution (content, SEO, community) | 8 | 2 | no | no | later — after a prototype exists | — |
| DM sales / appointment setting | 5 | 6 | no | partial (proposals are async DMs) | already has it — don't relearn | agency + real-estate jobs |
| Analytics / alerting (PostHog, Sentry) | 6 | 2 | Module M | no | later | — |
| Medical writing / lit reviews | 3 | 4 | no | no | never for money; research track only | — |
Top three now: directing + verifying, scoping, offer/copy — all three are exercised by the money path for free.
Explicitly not needed: another course, a CRM, setter training, the $99 Refactoring UI book (already vetoed in victor\DESIGN_COURSE.md L105).
Execution on approval
| # | Step | Output |
|---|---|---|
| 1 | Three agents in parallel, web-enabled. Every figure carries a source + year or reads unverified — never filled.
| three structured reports |
| 2 | Write the one deliverable doc — exhaustive, he skims chat but the file goes deep — and one line in the plans index. | C:\Users\BYJar\Desktop\workspace\plans\INCOME_1K_2026-08-22.mdC:\Users\BYJar\Desktop\workspace\plans\INDEX.md (+1 line, cap 40) |
| 3 | Inside that doc: the offer, the three portfolio pieces with URLs, the 60-second Loom script, the proposal template, the one-page spec template, retainer terms, the weekly cadence, the sourced skill table, rejected paths with numbers, the ## Timeline. | same file |
| 4 | Hour-zero rep, same day. He records the Loom (one take, sf-design.pages.dev) and sends proposal #1. Profile setup is prep, not progress. | one proposal timestamped today |
| 5 | Tonight is Saturday — the goals pass. The doc hands him one goal line to dictate; Signal writes nothing. Its first physical action: "open Upwork, filter last-24 h, send proposal #1". | his call on _briefings\agent\WEEKLY_GOALS.md |
Risks
| Risk | Mitigation |
|---|---|
| Upwork cold start — new profiles can go weeks at zero, and the first dollar must land before Sep 20 | 20 proposals in week 1, then 8/week; bid only on jobs <24 h old; Loom in every proposal; $80–150 48-hour wedge; Contra + Fiverr listings in parallel. Tripwire Sep 7: ≥20 sent and 0 replies → change the offer, not the channel — price the wedge at $49, broaden the niche to any small business, re-record the Loom. Tripwire Sep 14: still no paid job → take one ≤10 h contractor gig at whatever it pays (the 48–72 h-hire roles: VA, list-pulling, cold calling) purely to put a paid invoice on the board before Sep 20, then return to builds. A one-off, not a job. |
| Scope creep on fixed price | one-page spec approved day 1; one revision round; anything else is a new build |
| Collision with the course (D10 Aug 25, Module M after) and Year 4 (~Sep 20, 3 h commute) | bidding pinned to Sunday 21:00; builds are course reps; after Sep 20 drop to 1 build/month if the commute bites — retainers still compound |
| "Build the system, don't ship" — profile polishing, template perfecting | profile capped at 1 h; first proposal before the profile is finished; progress counted in proposals sent only |
| Payout friction — Upwork fee and hold, Payoneer to Egyptian bank | unverified as of 2026-08-22: 10% fee, ~5–10-day hold; Market agent confirms live |
| Reputation | cold market only — a thin month costs nothing in the warm market he is protecting |
Timeline
Aug22 Aug25 Aug29 Sep5 Sep7 Sep14 Sep19/20 Oct Nov
course ──D7───▶D10 end │ re-entry Year 4 starts (3h commute/day)
income 1st bid TODAY │ 20 proposals by Aug 29 ─┤ 8/wk ───────────────────────────────────▶
wedge $80–150 │ job 1 ───┤ job 2 ───┤ retainers 1–2 ──▶ 4–6 retainers = $1k/mo
tripwires │ target Sep5 Sep7: 0 replies→reprice Sep14: 0 paid→backstop gig
hard ────────────────────────────────────────────────────── FIRST $ IN BEFORE SEP 20 ──┤
hours 4h wk1 │ ~5h/wk: 2h bids + 3h build (builds = course reps)
$1k/mo recurring ≈ Oct–NovVerification
- Today: proposal #1 exists on Upwork with today's timestamp. Nothing else counts as progress.
- Weekly: the Sunday slot produced 8 proposals — count on the "My proposals" page, reported as a number in the evening brief, never rounded up.
- Deliverable doc: exists at
C:\Users\BYJar\Desktop\workspace\plans\INCOME_1K_2026-08-22.md; every rate has source + year or reads unverified; the skill table has a /10 score and a course-coverage column per skill; the Skeptic's disagreements are printed verbatim. - Aug 29: 20 proposals sent. Sep 7 / Sep 14: tripwires checked against reply count and paid count — acted on the same day, never deferred. Before Sep 20: first paid invoice visible in Payoneer — the hard line. Sep 22: two live URLs + two invoices. Nov 1: ≥4 retainer invoices paid in one month.
- Agreement: the plan is agreed only after he confirms the Timeline above.
Open Questions
Answer in chat. Each has a recommended default; silence takes the default.