WNAC / Wedge Needs A Copilot

What it is

WNAC (“Wedge Needs A Copilot”) is a static, local-first personal-finance dashboard. It keeps the parts of wnab that were worth keeping — clean, deduplicated transaction data and explainable AI categorization — and drops the envelope budgeting that made wnab a chore. The interface is a single screen built to feel like Copilot Money: one flex number for how much is actually free to spend, and a pay-down dial for debt.

How it’s built

Three plain files, no server, no cloud account: an import.py that normalizes and dedupes bank exports, an analyze.py whose “brain” is swappable (deterministic rules or an LLM), and a static dashboard.html. Everything runs on the local machine.

The privacy stance is built into the architecture: the method is public, the money is private. The code and design ship openly; the data, balances, and accounts never enter the repo and never leave the machine. That is why the screenshot above uses demo data, never real balances.

What worked

Dropping envelopes was the unlock. A single flex number answers the only question I ask daily — “can I spend this?” — without the assign-every-dollar ritual. The swappable AI brain means the dashboard degrades gracefully to plain rules when I don’t want a model in the loop, and the local-first design sidesteps the cold-start, subscription, and trust problems of cloud finance apps.

Lessons so far

• Keep the value, cut the ceremony. wnab proved the data pipeline and AI insight were worth having; WNAC keeps those and removes the part that never earned its upkeep.
• Privacy is an architecture decision. “Method public, money private” holds because the data never touches the repo or a server — the design enforces it, so the project can live in the open without leaking a number.

Status

Current. Active build, succeeding wnab.