More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Parallax One-Person Company

    1

    Build, launch and run your one-person business—from product decisions and development to marketing, delivery and ongoing improvement.

    $19

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Parallax One-Person Company

    Example session with this skill installed

    Founder request

    I am building a small digital-product business called Ledgerleaf. Use the complete company team to take this first product through a decision, actual implementation, finished marketing material and quality review. I want usable work, not a report about what I should ask someone else to do.

    The product is a browser-based retainer capacity planner for independent designers. It helps a designer compare available monthly client hours with committed retainers before taking another client. Build the working local product now. It must be usable without accounts, servers or paid services. No public deployment, outreach, real purchase or account access is authorized in this test. All people and businesses in these materials are fictional.

    Supplied customer notes

    • Interview A: an independent brand designer checks a spreadsheet before agreeing another monthly retainer. They want to see total committed hours and remaining room on one screen. They are interested in a tool, but did not agree to buy it.
    • Interview B: a freelance web designer keeps client allocations in calendar notes. They asked whether a planner could show an overbooked month before they say yes to a prospect. They did not name a price they would pay.
    • Interview C: a studio owner with eight staff wants payroll, timesheets, invoices and team permissions. They would not use a simple single-person planner for their whole studio.
    • Current alternative notes supplied by me: people can already do this in a spreadsheet. No current external competitor pricing, market-size estimate or conversion study has been supplied. Treat these as our internal fictional notes, not verified research.

    Product requirements

    • A complete locally usable web application with responsive styling and clear instructions. It must run through a normal static HTTP preview; no backend or external resource is needed.
    • Editable monthly total working hours, non-client work hours and reserved buffer hours. Available client capacity is total hours minus non-client work and buffer; do not quietly make impossible negative capacity look valid.
    • Add, edit and remove named client rows, each with committed monthly hours and a monthly fee in one chosen currency. Preserve names as literal text. Avoid labeling revenue as profit.
    • Show available client capacity, total committed client hours, remaining capacity or the overbooked amount, total monthly fee, and effective revenue per committed client hour. When committed hours are zero, explain that the effective hourly figure is not defined.
    • Clear validation for blank, negative, nonnumeric and inconsistent inputs; never silently turn missing values into an apparently valid zero.
    • Save and restore the user's scenario locally, with a clear reset control. Provide an export useful to the owner, such as a readable CSV or JSON file. Tell the owner where data is stored and what reset removes.
    • Include a labeled fictional starter scenario, but let the owner start a blank scenario. Use this starter: monthly total120h, non-client24h, buffer12h; Acorn36h/USD2160, Fieldline24h/USD1800, North Leaf18h/USD1260. A prospective client Moss would need12h forUSD900 per month; the user must be able to add that prospect themselves.
    • Do not add a contact form, payment checkout, analytics, account creation, automated customer messages or network calls.

    Business and launch work

    I can spend at most USD30 on the first validation attempt, but spending is not authorized in this test. My time budget is four hours per week after the prototype. My proposed selling price is USD9 one-time, which is a hypothesis, not proven willingness to pay.

    Make a product decision that accounts for the interview mismatch and the existing spreadsheet alternative. Preserve the small independent-designer audience and the functionality above. Produce a concise product/offer brief and a practical validation approach that fits the constraints.

    Ogilvy should produce finished launch assets: a landing-page draft for this actual product, three different launch posts and one opt-in follow-up email draft. Make the audience and next action concrete without inventing a public product link, enrolled users, testimonials or savings. Provide an acquisition experiment with a usable measurement record; do not contact anyone.

    Taleb should inspect the actual implementation, exercise meaningful cases, and check cost arithmetic and the marketing claims. Defects should go back to the appropriate owner for repair, followed by affected checks. Preserve which checks were actually run and what remains unverified.

    Save the company records and the complete deliverables in the assigned workspace. I should be able to hand the folder to another session and continue using the four departments.

    • Read your context and instructions
    • Compiled the parallax one-person company
    • Generated the data export

    From business idea to working product and launch materials

    Actual deliverables from the fictional Ledgerleaf local test. This is a product demonstration, not a customer result or sales claim. Ledgerleaf's proposed USD 9 price is a test hypothesis for that example product.

    1. Decide what to build

    The complete product and offer brief produced in the run:

    Ledgerleaf — product and offer brief v1

    Decision date: 2026-10-18. Owner: Munger. This is a bounded product decision, not evidence of demand.

    Decision and buyer

    Build the requested local retainer capacity planner for independent designers deciding whether another monthly retainer fits. Keep a single person, one month and USD as the chosen currency. The supplied fictional interviews A and B describe checking commitments before accepting work. Neither interview demonstrates willingness to pay. Interview C describes an eight-person studio seeking payroll, timesheets, invoices and permissions; those needs do not fit this product. Do not expand into studio operations.

    A spreadsheet already supports the arithmetic. Ledgerleaf's proposed reason to exist is a ready-made, focused screen with capacity validation, a visible overbooking warning, local persistence and export. Whether this convenience warrants a purchase is untested. USD 9 one-time is the founder's price hypothesis, not an approved public sale or a researched market price.

    Product behavior and acceptance

    • Runs as a static browser application with no backend, accounts, external libraries, external resources or application network calls. A static HTTP preview is the local delivery mechanism.
    • Total monthly hours minus non-client hours minus reserved buffer equals client capacity. All fields are required, finite and nonnegative. Combined non-client and buffer hours above total hours invalidates the plan instead of clamping capacity to zero.
    • Add, edit and remove named clients with monthly hours and monthly fees in USD. Names remain literal text. No HTML interpretation.
    • Display available client hours, committed hours, remaining or overbooked hours, monthly revenue and revenue per committed client hour. Zero committed hours produces an explicit undefined-rate explanation. Revenue is not profit.
    • Persist the raw scenario in this browser profile and origin; explain storage, limitations and reset behavior. Reset removes this application's saved scenario and returns to blank inputs. Export a readable, versioned JSON scenario, including raw invalid values if the owner needs a backup; identify validation status.
    • Default to the labeled fictional starter (120 / 24 / 12 hours; Acorn 36 h / USD 2,160; Fieldline 24 h / USD 1,800; North Leaf 18 h / USD 1,260). Provide a blank start. Do not include Moss by default; the owner can add the 12 h / USD 900 prospect.
    • Work on desktop and phone. Complete local checks and provide use/continuation instructions.

    First validation attempt

    Use a two-week, zero-spend proposed experiment after the prototype. Allow at most four founder hours per week: one hour preparing and selecting permitted channels; two hours of opt-in observation/replies; one hour recording and reviewing results. The USD 30 cap is unused; no spending, posting, outreach, account access, public link or sale is authorized here.

    Recruit independent designers who currently track at least one retainer, only after the founder separately authorizes a channel and contact action. Ask them to compare their existing method with entering a fictional month and adding a prospect; do not request private client data. Observe task completion, validation comprehension and whether they would choose the tool again. Ask a neutral USD 9 purchase-intent question after use and record reasons and objections. Purchase intent is not an order.

    Provisional decision rule: after five eligible completed observations, continue only if at least three complete the capacity task without help, at least three prefer this for that decision to their current method, and at least two independently describe a useful reason to choose it at the proposed price. Revise or stop if the main value is already met by their spreadsheet. This small convenience sample is directional, not statistical proof. Any checkout, actual sale or expansion requires a new authorized action.

    Ownership and next actions

    • Woz: implement the complete static application and document local use.
    • Ogilvy: finished landing-page draft, three distinct launch posts, one opt-in follow-up email draft, experiment plan and empty measurement record. Use actual capabilities and clearly state local prototype status.
    • Taleb: inspect the delivered version, run relevant behavior checks, recompute scenario and budget arithmetic, inspect marketing claims, return defects to their owners and recheck repairs.
    • Munger: reconcile the checkpoint and use actual future observations to choose the next bounded iteration.

    2. Build the working product

    Download the complete generated planner · Inspect the application files

    The delivered HTML, CSS and JavaScript application lets the owner edit monthly capacity and client rows, validates missing and inconsistent values, saves the scenario in the browser and exports readable JSON. It includes desktop and phone layouts.

    Actual scenarioAvailable hoursCommitted hoursResultMonthly fees
    Supplied fictional starter84786 hours remainingUSD 5,220
    After adding Moss at 12 hours / USD 90084906 hours over capacityUSD 6,120

    The second gallery image shows the actual delivered application. The downloadable folder runs through a local static preview; no backend, account or external application dependency is required.

    3. Prepare the launch

    The complete landing-page copy from the run follows. Its reply CTA belongs to the fictional, unpublished recruitment draft.

    See whether your next retainer fits.

    Ledgerleaf is a monthly capacity planner for independent designers. Put your available client hours and existing retainers on one screen, then add a prospect to see how the month changes.

    Try the local prototype with fictional clients. Reply

    “planner” to the post or message that brought you here to request an opt-in test.

    Give the month a realistic limit

    Start with your total working hours. Set aside time for non-client work and a buffer. Ledgerleaf shows the hours left for client commitments.

    Add each retainer's name, monthly hours and fee in USD. Edit or remove a row as your plans change. The totals show:

    • Available client capacity and hours already committed.
    • Hours remaining, or how far the plan is overbooked.
    • Total monthly client fees and revenue per committed client hour.

    If required values are missing or invalid, the planner asks you to correct them. Non-client work and buffer cannot exceed the month's total hours. When there are no committed client hours, the hourly revenue figure is undefined.

    Try this fictional month

    Monthly planHours
    Total working time120
    Non-client work24
    Reserved buffer12
    Available for clients84
    Existing retainers78
    Remaining capacity6

    The starter's three fictional clients total USD 5,220 in monthly fees. Add a prospect needing 12 hours at USD 900 a month yourself. Committed time becomes 90 hours, putting the plan

    6 hours over capacity. Monthly fees become USD 6,120.

    The fee totals describe the scenario you entered. They are revenue figures before expenses and taxes, not profit or a prediction of what you will earn.

    Keep a local scenario

    Use the labeled fictional starter or begin with blank inputs. The prototype runs in a browser through a local static preview, without an account or backend. The app makes no external calls.

    Your scenario saves in this browser profile for the preview address. It does not sync to another device or browser. Clearing browser storage, using a temporary private session or changing the preview address can make that saved scenario unavailable.

    Export a readable JSON snapshot when you want a separate copy. Export is not an automatic backup or a promise of import support. Reset clears this app's saved scenario and returns the planner to blank inputs. It does not delete exported files.

    One designer, one month

    Ledgerleaf covers monthly retainer hours and fees in USD. It does not include payroll, invoices, timesheets, team permissions or calendar scheduling. Your estimates still determine whether the plan is useful.

    A spreadsheet can already do this arithmetic. The question for this prototype is whether a dedicated screen makes this particular decision more convenient for you.

    Help test that question

    We are testing a possible

    USD 9 one-time price. That is a pricing hypothesis. The current local prototype test is free, and there is no checkout or purchase available here.

    If you are an independent designer with at least one retainer, reply

    “planner” where this page was shared. You can take part in a short, opt-in test using the fictional month, compare it with your current method and tell us what you would keep or change. No private client information is needed.

    The run also delivered three complete launch posts, an opt-in follow-up email draft, an acquisition experiment and blank measurement records. No outreach was sent in this test.

    4. Review, repair and improve

    Taleb inspected the real implementation and marketing. It found a banner that could say the scenario was saved after a blank reset or a storage failure. Woz changed that banner to point to the actual save status. Taleb rechecked the affected browser behavior and controlled storage-failure paths.

    The reviewed product v1.0.1 had no unresolved blocking finding in the inspected scope. The engineering suite passed 15 checks and a separate Taleb worker passed 18 checks. The run covered calculations, input validation, saved scenarios, actual JSON export and responsive layouts. These checks do not establish compatibility with every host or device.

    Read the actual review · Inspect the company record

    Continue the company

    The run saved department owners, decisions, task and review states, and next actions. A separate fresh-session test then corrected the product's example values and the affected marketing while preserving old reviews and separately saved user scenarios.

    In this run, Munger and Woz were passes by the main worker; Ogilvy and Taleb used separate workers. No four simultaneous background processes or real customer outcome is claimed.

    Open all selected actual outputs and the original request.

    ledgerleaf-example.zip

    CSV · data export

    Generated

    Example file from a real run - the skill writes it into your workspace.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Parallax One-Person Company helps you turn a business goal into working products, marketing materials and customer delivery. You set the direction and requirements. Your AI team plans the work, carries it out, checks the results and keeps the records needed to continue.

    Build, launch and run your business

    • Choose what to build: investigate customer needs, shape your offer and decide what deserves your time.
    • Produce the work: build software, websites, integrations and other deliverables with your assistant's tools.
    • Bring it to customers: prepare product pages, launch posts and customer-facing material, and carry out authorized marketing actions.
    • Deliver and improve: review actual outputs, fix problems, manage client changes and turn feedback into the next improvement.

    Choose your team in v1.3

    Team 1 — Business and growth. The default team: Munger handles business and product decisions, Woz engineering and delivery, Ogilvy marketing and customer work, and Taleb quality and risk review. A useful starting point for offers, commercial choices and growth.

    Team 2 — Product and delivery. Elon coordinates tasks and dependencies. Jobs handles product and design, Linus implementation, Turing verification and testing, and Bezos customer success and delivery. A useful starting point for coordinated software development and ongoing customer work.

    Say “Switch to Team 2” to change the whole team, or “Switch to Team 1” to switch back. You can also select colleagues from either team and save a custom team. Address one colleague directly for a focused task, or invite someone for a one-task consultation.

    Continue the same company

    Team changes preserve tasks, decisions, work files and review history. Unfinished work gets an owner in the selected team. You remain the human founder; routine execution and recordkeeping belong to the AI team.

    Start with your goal, relevant materials and constraints. Ask the team to produce the deliverables, check them and save the company records. File access supports continuation; text-only sessions provide a handoff to reuse.

    Separate workers depend on your host; otherwise roles run as labeled passes. AI usage, hosting and paid services are separate. The Ledgerleaf example shows an actual earlier local run with the first team.

    This Skill does not include model accounts, a hosted company dashboard, the ParallaxAI gateway runtime, a server or always-on employees.
    Actual code execution, rendering, integration, deployment and external research depend on the host's available tools and access.
    No sale, traffic, conversion, product-market fit or income is guaranteed. Findings depend on the supplied evidence and the checks actually performed.
    The Skill retains the requested business responsibilities when a tool is unavailable; a plan or mockup is not reported as a finished working product.
    Reviews are practical checks of the actual artifact. They are not professional certification or an independent audit when performed in the same context.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 26 days ago

    Permissions

    Read Files
    Write Files
    Terminal / Shell
    Browser
    Network Access

    File Scopes

    client-delivery/**
    parallax-company/**

    Uses the project files, terminal, browser and independent workers available in your AI host for the work you request. Business records and deliverables are saved in an authorized workspace. Publishing, sending messages, spending, accounts and destructive changes follow the user's scope and host confirmation rules. Loading the Skill alone does not authorize those actions.

    Listed1 month ago
    Updated26 days ago

    What's inside

    Frequently Asked Questions

    Use it to develop a product, launch a service, improve an offer or manage client work. Give it your business goal, relevant materials and constraints. It connects decisions, implementation, marketing and review, produces the work its tools permit and keeps records for the next session.

    Yes. In v1.3, Team 1 is the default: Munger, Woz, Ogilvy and Taleb. Switch the whole team to Elon, Jobs, Linus, Turing and Bezos, or save a custom team using colleagues from either preset. You can also address one colleague directly without changing the saved team.

    Yes. The engineering colleague creates and checks working deliverables using your assistant's tools. The marketing colleague produces finished copy, launch assets and relevant research. The team reviews actual outputs and repairs defects. A plan or role discussion does not count as a requested working product.

    You provide the business goal, requirements and constraints, and retain final business authority. The AI team handles routine task breakdown, execution, handoffs, checks and records. Your assistant uses separate workers when supported, or labeled role passes otherwise. Names and roles do not create always-on employees or additional permissions.

    With file tools, Parallax saves company records and reads them when you resume. In v1.3, switching teams preserves tasks, decisions, artifacts and review history while assigning unfinished work to new owners. Your saved team remains active. Without file access, reuse the continuation handoff in the next conversation.

    The upgrade preserves the original checkpoint and work files, carries forward the actual scope, tasks and approval evidence, and records their mapping to the full team. An old approval does not automatically accept a changed version or new scope.

    Those actions remain available work for the relevant department when you request and authorize them and the host permits them. The Skill does not treat installation as permission to contact people, access accounts, spend or deploy.

    No. The one-time price buys the reusable Skill. AI usage, hosting, domains and optional external services remain separate.

    New in Business & Operations