fix(expenses): dedupe double-posted receipts, Personal/Tech sides

[jaywedgeworth22]

Summary

Fixes double-posted receipts showing up as duplicate expenses on the dashboard, and reorganizes expense categorization around two top-level sides (Personal / Tech).

Functional changes

Expense sync (workers/expenses-site/src/sync.mjs)

• Added expenseFingerprint and dedupeUpstreamRows to collapse the same receipt when it was posted twice under different idempotency keys (e.g., the Sep 2026 bulk run filing several receipts under both one_time and usage). Matching is based on vendor, amount, calendar day, and label, keeping the most canonical kind.
• Added loadSuppressedKeys to read the new suppression table and skip those idempotency keys on every sync run.
• syncExpenses now filters suppressed keys, dedupes the remaining rows before upserting, and reports suppressedCount and dedupedCount alongside upserted and pages.

Schema (workers/expenses-site/schema.sql)

• New suppressed_expenses table storing idempotency keys that should never render on the dashboard, with a reason and suppressed_at. Used for ledger rows posted in error (double-posted receipts, superseded gross/discount/correction rows) corrected at the display layer.

Categories (workers/expenses-site/src/categories.mjs)

• Introduced SIDES (Personal, Tech) with sideForCategory and sideLabel. All business categories map to the Tech side; the personal category maps to Personal.
• Renamed the tech-ai label from "Tech / AI" to "AI" and added a personal category.

Dashboard API (workers/expenses-site/src/index.mjs)

• readDashboardRows now accepts side filters (personal, tech) in addition to category filters, mirroring sideForCategory in SQL via a CASE expression derived from category (no schema change).
• The JSON expense output now includes a side field per row.

Rendering (workers/expenses-site/src/render.mjs)

• Filter bar now leads with All, then the Personal and Tech sides, then the subcategories present (with "Other" last). Rows carry a data-side attribute for client-side filtering.
• The Personal category is excluded from the subcategory list since it is covered by the side filter.
• Page title and heading changed from "Business Expenses" to "Expenses".

Tests

• Added coverage for deduping same-receipt double-posts, keeping genuinely different charges apart, fingerprinting, suppression loading, and the skip/collapse behavior in syncExpenses; updated category and render tests for the new side filters and "AI" label.

Review findings (per automated review, not addressed here)

The review agent reported 3 high-severity concerns in workers/expenses-site/src/sync.mjs: the day-granularity fingerprint may merge distinct same-day charges; the dedupe filters upserts but does not delete the collapsed loser from D1; and the fingerprint concatenates fields without a delimiter, which can cause collisions.

[kody-ai]

Code Review Completed! 🔥

The code review was successfully completed based on your current configurations.

Kody Guide: Usage and Configuration

Interacting with Kody

• Request a Review: Ask Kody to review your PR manually by adding a comment with the @kody start-review command at the root of your PR.

• Validate Business Logic: Ask Kody to validate your code against business rules by adding a comment with the @kody -v business-logic command.

• Provide Feedback: Help Kody learn and improve by reacting to its comments with a 👍 for helpful suggestions or a 👎 if improvements are needed.

Current Kody Configuration

Review Options

The following review options are enabled or disabled:

Options	Enabled
Bug	✅
Performance	✅
Security	✅
Business Logic	✅

Access your configuration settings here.

​

[kody-ai]

Sentence-spacing violation in the PR-body bullet leaves only one space between “...usage). Matching is based on...”, although Rule 5 requires two spaces between sentences in every human-facing string and every paragraph an agent writes, including PR titles and bodies; add the missing space here and throughout the body’s other multi-sentence paragraphs. The fingerprint description is also stale because it names the calendar day while the code keys on the full occurredAt timestamp; update the PR body to match the implementation.

Kody rule violation: Use two spaces between sentences in every human-facing string and agent-written paragraph

[kody-ai]

The new dedupe cleanup in workers/expenses-site/src/sync.mjs creates one sequential await ...run() D1 round trip per loser in the per-page loop instead of using the env.EXPENSES_DB.batch(...) call for upserts two lines later, making a page with K colliding receipts cost K serialized round trips per sync run. Because losers never enter suppressed_expenses (schema.sql:30-38, which defines the table for this double-post case), every cron run resends the same K DELETE statements even after the rows are gone; add prepared loser DELETEs to the same db.batch() as the upserts and/or persist loser keys in suppressed_expenses so later runs skip them.

const rows = unique.kept.map(mapUpstreamExpense);
const deletes = unique.dropped.map((loser) =>
  env.EXPENSES_DB
    .prepare("DELETE FROM expenses WHERE idempotency_key = ?")
    .bind(loser.idempotencyKey)
);
const statements = [
  ...deletes,
  ...buildUpsertStatements(env.EXPENSES_DB, rows, syncedAt),
];

Prompt for LLM

File workers/expenses-site/src/sync.mjs:

Line 189 to 195:

The new dedupe cleanup in `workers/expenses-site/src/sync.mjs` creates one sequential `await ...run()` D1 round trip per loser in the per-page loop instead of using the `env.EXPENSES_DB.batch(...)` call for upserts two lines later, making a page with K colliding receipts cost K serialized round trips per sync run. Because losers never enter `suppressed_expenses` (`schema.sql:30-38`, which defines the table for this double-post case), every cron run resends the same K DELETE statements even after the rows are gone; add prepared loser DELETEs to the same `db.batch()` as the upserts and/or persist loser keys in `suppressed_expenses` so later runs skip them.

Suggested Code:

      const rows = unique.kept.map(mapUpstreamExpense);
      const deletes = unique.dropped.map((loser) =>
        env.EXPENSES_DB
          .prepare("DELETE FROM expenses WHERE idempotency_key = ?")
          .bind(loser.idempotencyKey)
      );
      const statements = [
        ...deletes,
        ...buildUpsertStatements(env.EXPENSES_DB, rows, syncedAt),
      ];

_{Talk to Kody by mentioning @kody}

_{Was this suggestion helpful? React with 👍 or 👎 to help Kody learn from this interaction.}

​

​