Should I vibe code
Family portfolio tracker with live valuations, UK planning tools, and AI reviews
Tracking your portfolio is fine. The moment it says “you should”, you have built advice.
?
Their verdict, the AI credits price and the build-time estimate come from their entry, MIT-licensed. Checked 2026-08-03.
?
Our verdict, the regret score and everything below it. Editorial and unsponsored — nobody can pay to be moved.
The honest answer
why the verdict is what it is
Pulling valuations and charting them is a good project. Keep it descriptive: a homemade tool that makes planning recommendations is a tool you will trust more than it deserves.
What actually breaks
not "if". the specific failures.
- The line between tracking and advice, which is a regulatory boundary rather than a product decision
- Price data licensing, where market data feeds have redistribution terms and 'free API' rarely means free to show other people
- Corporate actions — splits, dividends, mergers — which silently make a historical position wrong if unhandled
- Tax logic, which is jurisdiction-specific, changes annually, and is the part people most want automated
- Family access, so a spouse or an adviser can see holdings, which turns a personal tool into a multi-user one holding financial data
The tracker shows a healthy return. It has not handled the three-for-one split from March, so the historical cost basis is still per pre-split share while the current price is post-split — the position looks like it tripled. You make a decision based on it. The correction is arithmetic and takes a minute; the trade you already placed on the strength of the wrong number is not undone by fixing the display.
Is that you?
the verdict is a default, not a law
- It tracks and reports, and offers no suggestions about what to do
- Positions are entered manually and prices are for your own display only
- It is a single-user tool and you know the limits of its arithmetic
- It recommends actions, allocations or rebalancing to anyone
- Corporate actions are unhandled but historical returns are displayed
- You are redistributing market data to other people
- It calculates tax positions people would file from
If you build it anyway
the checklist, then the prompt that enforces it
- Draw the advice line explicitly and stay on the tracking side. Describing a portfolio is not the same activity as recommending changes to it, and the difference is regulatory.
- Read the licence on any market data source before displaying it to anyone but yourself. Redistribution is usually the restricted part.
- Handle splits, dividends and mergers or refuse to display historical returns. Half-handled corporate actions produce confidently wrong numbers.
- Label tax figures as estimates, cite the rule you applied, and tell people to check with an accountant.
- If anyone else can see it, that is multi-user financial data — authenticate properly and decide who sees what.
- Show the data's timestamp and source next to every valuation.
Before you build a portfolio tracker, apply these and push back if I ask you to break them. 1. Ask me whether this will suggest what to do — buy, sell, rebalance, allocate. If yes, tell me that is the line between tracking and advice, that the line is regulatory in most jurisdictions, and recommend staying on the tracking side. 2. Check the licence terms of any market data source before displaying prices to anyone other than me. Explain that redistribution is normally the restricted right, and that a free API is often personal-use only. 3. Handle corporate actions — splits, reverse splits, dividends, mergers, spin-offs — before displaying any historical return. If they are not handled, refuse to show historical performance and say why. 4. Represent money and share quantities as integers or decimals, never floats, and store the currency with every amount. 5. Label all tax calculations as estimates, state which rule and which tax year was applied, and tell the user to verify with an accountant. Do not produce figures that look filing-ready. 6. Show the source and timestamp of every valuation next to the number. 7. If family members or an adviser will have access, treat this as multi-user financial data: real authentication, per-user visibility, and an audit of who saw what. 8. Never store brokerage credentials. Manual entry or file import only. 9. Out of scope unless I ask: automated trading, performance benchmarking against indices, retirement projections.
That one keeps you out of trouble. For the prompt that actually builds it, canivibecodeit.com has one.
their build prompt ↗Or don’t build it
the boring option, and the way back out
The corporate actions and licensed price data are the parts worth paying for, because both produce silently wrong numbers when done casually. A manual tracker that reports positions honestly and stops short of advice is a sound personal project.
Keep positions, transactions and corporate actions as exportable rows with dates and currencies — the transaction history is the asset, and prices can always be re-fetched. Anything the tool computes should be reproducible from that history rather than stored as a result.
Open-source wealth management software with portfolio analytics
Open-source desktop portfolio tracker with detailed performance analysis
Questions
Where exactly is the line between tracking and advice?
Roughly: describing what you hold and what it is worth is tracking; suggesting what you should do about it is advice. Rebalancing prompts, allocation targets and 'you are overweight in X' sit on the advice side in many jurisdictions. The specifics vary, which is itself a reason for a personal project to stay well clear of the boundary.
Why do corporate actions cause such confident errors?
Because they change the meaning of historical data without changing its shape. After a split, old prices and quantities are still valid numbers — they are simply in different units from current ones. Nothing looks malformed, so no validation catches it, and the resulting return figure is plausible enough to act on.
Every week, someone ships something they shouldn’t have.
New verdicts, the worst thing that landed in the trap, and the occasional incident report. No other email, ever.
last reviewed 2026-08-03 · verdict is editorial and unsponsored · shared entry data from canivibecodeit under MIT · not legal advice