BrewCalc tracks shared coffee costs in a workplace setting. When colleagues share a coffee machine and take turns buying beans, it becomes difficult to track who owes what. BrewCalc solves this by recording coffee bag purchases and individual drink consumption, then computing a fair cost split every time a bag is finished.
The tool calculates a per-drink cost (bag cost divided by the drinks served from that bag), multiplies by each person's consumption, and offsets against what they paid for bags. The result is a simple balance: credit (overpaid) or debt (underpaid). Settlements close automatically when a bag is emptied — either manually or when a new bag is opened.
BrewCalc uses the default lab.johlem.net stack:
api.php endpoint handling all routes via query parameter dispatch.coffee.db) is created automatically on first use.| Table | Purpose | Key Fields |
|---|---|---|
groups | Workspace per barista | name, created_at |
users | Login accounts | username, pin_hash, role, group_id, person_id |
people | Coffee drinkers (group members) | group_id, name, emoji |
coffee_bags | Bag purchases | buyer_id, total_cost, name, is_empty, emptied_at |
drink_logs | Individual drinks (attributed to a bag) | person_id, bag_id, created_at |
settlements | Closed bag settlements | group_id, bag_id, mode='bag', closed |
credits | Manual balance adjustments / cash transfers | person_id, amount, note |
bag_ratings | Per-person bag ratings + tasting notes | bag_id, person_id, stars, note |
bag_suggestions / bag_suggestion_votes | Proposals for the next bag + upvotes | name, status, picked_bag_id |
audit_logs | Append-only activity log for baristas | actor, action, summary, created_at |
A settlement is generated each time a bag is closed. The algorithm scopes to that bag's lifetime and to drinks attributed to it.
total_cost is divided by the drinks whose bag_id equals the bag — so cost-per-drink stays stable even across month boundaries.cost share = drinks_from_bag × cpd, paid = bags bought in the window, credits = credits booked in the window.cost share − paid − credits. Positive = owes money, negative = is owed money.Each emptied bag produces a yield sample: cups = COUNT(drink_logs WHERE bag_id = b.id). The server averages this per brand, falling back to group-wide average. For the active bag the dashboard shows:
max(0, expectedCups − drinksIn).ceil(cupsLeft / dailyRate), where dailyRate is the group's last-14-day average.escHtml() to prevent XSS.X-Content-Type-Options, X-Frame-Options, Referrer-Policy.Security limitations tracked for a public launch are documented in PHASE2.md at the repo root.
BrewCalc stores only the minimum data necessary: usernames (derived from group name), names, emoji avatars, drink counts, bag costs, and optional tasting notes. No email addresses are collected at this phase.
BrewCalc supports full JSON export and import via the Data tab:
PHASE2.md).bag_id attribution and is excluded from per-bag settlements. The FIFO logic prevents this in normal use but imports can produce it.localStorage; it does not affect stored amounts.