Document that both backend/package.json and frontend/package.json
must be updated before tagging a release, since the About modal
reads versions from these files.
Fixes#96
- nginx.conf converted to template processed by envsubst at container start
- BACKEND_URL env var (default: backend:3000) replaces hardcoded container name
- Docker DNS resolver used for dynamic upstream resolution
- Dockerfile copies nginx.conf as template to /etc/nginx/templates/
This prevents frontend breakage when users customize container names
in their docker-compose.yml.
- getDataDir() now detects monorepo by checking for ../docker-compose.yml
- DATA_DIR env var removed from .env and .env.example (no longer needed for local dev)
- Docker compose files explicitly set DATA_DIR=/app/data for containers
- Updated tests for monorepo detection logic
Add DATA_DIR env var support to configure the data directory path.
All hardcoded resolve(cwd, 'data') paths now use a central getDataDir()
function from db-utils.ts that checks DATA_DIR first, falling back to
resolve(cwd, 'data').
This prevents local dev (cd backend && npm run dev) from creating a
separate backend/data/ directory instead of using the root data/ folder.
Changes:
- Add getDataDir() to db-utils.ts as single source of truth
- Update all 8 source files that reference the data directory
- Add dotenv fallback to ../.env for local dev from backend/
- Add DATA_DIR documentation to .env.example
- Add 7 new tests for getDataDir and getDbPaths with DATA_DIR
- 493 tests pass, TypeScript clean
Two bugs in the backend medications route:
1. Planner /medications/usage had the same +1 phantom consumption bug
that was fixed in the frontend (PR #109). After a stock correction,
effectiveStart was set to max(blisterStart, correctionCutoff) instead
of correctionCutoff + period, causing 1 dose to be immediately
counted as consumed.
2. PUT /medications/:id did not reset stockAdjustment when stock fields
(packCount, blistersPerPack, pillsPerBlister, looseTablets) changed.
If a user edited stock values to correct their inventory, the old
stockAdjustment offset was preserved, resulting in wrong totals.
Added 4 tests covering both scenarios.
Extract DB utility functions (buildDbUrl, getDbPaths, ensureDataDirectory,
runAlterMigrations, etc.) from client.ts into db-utils.ts.
client.ts contained top-level initialization code (ensureDataDirectory,
createClient) that ran on every import. database.test.ts imported utility
functions from client.ts, which triggered the initialization as a side
effect — creating backend/data/ with a .write-test file and
medassist-ng.db every time tests ran.
Now database.test.ts imports from db-utils.ts (side-effect-free), and
client.ts re-exports everything for backward compatibility.
test.yml: Use dorny/paths-filter to detect changed paths. Backend
tests only run when backend/**, biome.json, or the workflow itself
changes. Frontend build only runs when frontend/**, biome.json, or
the workflow changes. Jobs skipped via job-level 'if:' are treated
as passed by GitHub required checks.
codeql.yml: Only run on push/PR when JS/TS source files, package
files, or CodeQL config changes. Weekly schedule and manual dispatch
remain unfiltered.
Create dedicated GitHub Copilot agent for release management with
4 tasks: branch/PR workflow, version determination, release execution,
and release notes writing.
Move release-specific instructions (workflow, release notes format,
breaking changes) from copilot-instructions.md to the agent file
to keep concerns separated.
After correcting medication stock, the coverage calculation immediately
counted 1 dose as consumed (due to +1 in occurrences formula), which
neutralized small corrections like +1 pill.
Fix: start consumption counting from stockCorrectionCutoff + period
(the next scheduled dose) instead of from the correction time itself.
Added 3 frontend tests for stock correction scenarios and 6 backend
e2e tests for the PATCH /medications/:id/stock-adjustment endpoint.
The groupedSchedule useMemo used slice(0, 2000) to limit events. With daily
medications having start dates far in the past, thousands of past events would
fill all 2000 slots, pushing today and future events completely out of the
display. This caused the past schedule to only show weekly medications (fewer
events) while daily medications appeared missing.
Replace the fixed count limit with a time-based window: only past events
within the scheduleDays window (30/90/180 days) are included. All today and
future events are always included regardless.
Coverage calculations are not affected as they use schedule.events directly.
The takenBy field is a string[]. Empty arrays [] are truthy in JavaScript,
causing d.takenBy ? [...] patterns to generate dose IDs with trailing
hyphens (e.g., '5-0-173...-') instead of base IDs ('5-0-173...').
This mismatch between ID generation and computeMissedPastDoseIds (which
correctly uses .length > 0) caused doses to always appear as missed.
Changes:
- Add expandDoseIds() helper function using correct .length > 0 check
- Replace 8 buggy inline patterns in DashboardPage.tsx
- Refactor SchedulePage.tsx to use shared expandDoseIds()
- Add backend startup repair to strip trailing hyphens from existing IDs
- Add 12 new tests (6 frontend + 6 backend)
Add repairOrphanedDoseIds() function that runs during app startup
(after ALTER migrations) to fix dose tracking entries that became
invalid when medication schedules were changed before PR #103.
The function:
- Generates valid schedule dates for each medication's current intakes
- Detects dose_tracking entries whose dateOnlyMs doesn't match any
valid schedule date
- Remaps orphaned doses to the nearest valid schedule date within
half the intake interval
- Preserves person suffixes in dose IDs
- Is idempotent (safe to run on every startup)
This complements PR #103 which only migrates dose IDs on future edits.
The startup repair fixes existing broken data in production databases.
Includes 8 tests covering: valid doses unchanged, 1-day shift repair,
person suffix preservation, out-of-range detection, idempotency,
multi-medication handling, and legacy format fallback.
When a medication's start date or interval changes, the generated dose
IDs shift (dateOnlyMs values change). Previously, doses marked as taken
under the old schedule were orphaned — they no longer matched the new
schedule's dose IDs, causing them to appear as missed.
Now the PUT /medications/:id endpoint:
1. Parses old intakes from the existing medication row
2. Detects which intake indices had schedule changes
3. Maps old dateOnlyMs values to the nearest new dateOnlyMs
4. Updates dose_tracking entries with the migrated IDs
5. Preserves person suffixes (e.g. -Alice) during migration
Also fixes the start-date cleanup to use date-only comparison,
preventing doses on the start date from being incorrectly deleted
when the start time is after midnight.
Adds 4 integration tests covering weekly day shift, person suffix
preservation, time-only changes, and interval changes.
Allows manual triggering of the badge update workflow, useful when
the ANSI fix or other workflow-only changes need to take effect
without waiting for source code changes.
Vitest 4 outputs ANSI color codes in test results, which caused the
grep regex to fail when extracting test counts. The badge workflow
silently skipped the update, leaving stale counts in the README.
Add a sed pass to strip ANSI escape sequences before parsing.
- Add 'Include consumption from today until start date' checkbox to planner
- When checked, usage calculation starts from today instead of max(today, startDate)
- Persist checkbox state in localStorage per user
- Add i18n translations (EN + DE)
- Update planner tests to use dynamic future dates
- Remove broken isDoseFromPreviousSchedule that falsely dismissed all past doses
after any medication edit (compared dateOnlyMs < updatedAt incorrectly)
- Fix takenBy normalization in AppContext: event.takenBy (string|null) was passed
through as-is via || operator instead of being properly converted to string[]
- Fix DashboardPage: 5 locations treated dose.takenBy as single string instead of
iterating the array, causing per-person dose tracking to silently fail
- Extract isDoseDismissed and computeMissedPastDoseIds as pure testable functions
from AppContext.tsx into utils/schedule.ts
- Update SharedSchedule.tsx to use shared isDoseDismissed from utils
- Add 22 regression tests covering isDoseDismissed, computeMissedPastDoseIds,
and full dose-tracking-survives-medication-edit workflows
- Add 'fix bugs, don't test around them' rule to copilot instructions
* Initial plan
* Add Playwright E2E testing infrastructure
- Add @playwright/test dependency
- Create playwright.config.ts with best practices configuration
- Create e2e test structure with fixtures and auth setup
- Add E2E tests for auth, dashboard, medications, and settings pages
- Add npm scripts for running E2E tests
- Update .gitignore for Playwright artifacts
- Add E2E test job to CI workflow
- Update vite.config.ts to support BACKEND_URL env variable
- Update biome.json to include e2e files in linting
Co-authored-by: DanielVolz <3275994+DanielVolz@users.noreply.github.com>
* Remove waitForTimeout anti-pattern from E2E tests
Replace hard-coded timeouts with proper Playwright waiting strategies:
- Use waitForLoadState('networkidle') for page load
- Use element.waitFor() for dynamic elements
- Use expect assertions for state verification
Co-authored-by: DanielVolz <3275994+DanielVolz@users.noreply.github.com>
* Remove E2E tests from CI workflow
E2E tests will only be run locally as requested.
Co-authored-by: DanielVolz <3275994+DanielVolz@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: DanielVolz <3275994+DanielVolz@users.noreply.github.com>
Co-authored-by: Daniel Volz <mail@danielvolz.org>
- Add Array.isArray() checks before calling .map() on dose.takenBy
- Fixes TypeError: dose.takenBy.map is not a function
- Affects AppContext missedPastDoseIds calculation
- Affects SchedulePage dose ID generation (3 locations)
This hotfix prevents the app from crashing when dose.takenBy
is null, undefined, or any non-array value.
## Package Type Feature
- Add 'blister' and 'bottle' package types for medications
- Bottle type uses totalPills for capacity and looseTablets for current stock
- Blister type continues to use packCount/blistersPerPack/pillsPerBlister
- Add doseUnit field for flexible dosing (mg, ml, IU, etc.)
- Full UI support in medication form and detail modal
## Per-Intake TakenBy
- Move takenBy from medication level to individual intakes
- Each intake schedule can now be assigned to a different person
- Update scheduler-utils to handle per-intake takenBy
- Update SharedSchedule to filter by per-intake takenBy
- Backward compatible with existing medication data
## UI Improvements
- Add PasswordInput component with show/hide toggle
- Centralize stockThresholds in AppContext for consistent status display
- Fix SharedSchedule sync issues with per-intake takenBy
- Improve mobile editing experience
## Technical
- Add migrations 0004 and 0005 for schema changes
- Update all relevant tests (1064 tests passing)
- Maintain backward compatibility with ALTER migrations
- Add updatedAt field to share API response
- Add isDoseFromPreviousSchedule check in SharedSchedule
- Don't count doses scheduled before medication update as missed
- Syncs SharedSchedule behavior with main app's AppContext logic
- Hide Reorder Reminder card when reminders are enabled (avoids redundancy with Reminder Bar)
- Show all low stock medications in Reminder Bar instead of just the next one
- Rename 'Reorder' to 'Refill' throughout the app
- Make medication names clickable in Refill Reminder card (opens detail modal)
- Add daysLeft display for each low stock medication
- Update translations (EN + DE)
* feat: add account deletion feature
- Add DELETE /auth/me endpoint to delete user account and all data
- Add deleteAccount() method to AuthContext
- Add Delete Account button with confirmation modal in UserProfile
- Add danger zone styling (.btn-danger, .profile-danger-zone)
- Add i18n translations for EN and DE
- Add backend tests for account deletion endpoint
- Add timeout settings to frontend vitest.config.ts
- Reduce CI timeout for frontend tests (10min -> 5min)
* fix: improve delete account section layout
- Make profile modal scrollable with max-height
- Add proper horizontal margin to danger zone
- Align delete section with form content
* fix: use ConfirmModal component for delete account dialog
- Replace inline modal with existing ConfirmModal component
- Ensures consistent button styling across all modals
- Add UI consistency rule to AGENTS.md and copilot-instructions.md
* fix: consistent styling for delete account section
- Remove warning text (users know what delete means)
- Remove border-bottom from danger zone title (section has border-top)
- Update copilot-instructions and AGENTS.md with stricter UI consistency rules
- Remove unused deleteAccountHint i18n keys
* chore: remove pre-push test hook (CI handles tests)
Tests were running twice - in pre-push hook and GitHub CI.
Removing local pre-push tests since CI provides authoritative test results.
Use 'npm test' manually before pushing if you want local feedback.
- Add test:run script to frontend package.json (consistent with backend)
- Use npm run test:run instead of npm run test -- --run
- Add timeout-minutes to prevent infinite hangs
- Use date-only timestamp instead of full timestamp for dose ID generation
- Ensures changing intake times doesn't invalidate past dose tracking
- IDs are now immune to time configuration changes
- Consolidate duplicate date formatting utilities
- Use shared formatters across backend and frontend
- Clean up test mocks to use consistent test data
- Remove redundant formatting functions
When medication intake times change, dose IDs change (they include
timestamps). Previously, this caused all past doses to appear as
'missed' because the old 'taken' markers no longer matched.
Now doses are only counted as 'missed' if they were scheduled AFTER
the medication's last update (updatedAt). This means:
- Legitimately missed doses still show as missed (e.g., yesterday's
dose not taken)
- Doses from before a schedule change are NOT counted as missed
(they were from a previous schedule configuration)
Changes:
- AppContext: Add isDoseFromPreviousSchedule helper
- SchedulePage: Use context's missedPastDoseIds instead of local calc
- Update tests to include missedPastDoseIds in mocks
Creates a workflow that triggers when a release is published and
automatically updates package.json versions in backend/ and frontend/
to match the release tag version.
The release script created tag v1.6.0 but did not update the version
numbers in package.json files. This fix ensures the About modal
displays the correct version.
Update all visible text from 'MedAssist' to 'MedAssist-ng':
- Auth page titles (login, register)
- Loading/error/initializing states
- SharedSchedule page (loading, expired, error, footer)
- AboutModal fallback text
- i18n strings for export file validation (EN/DE)
- Related test expectations
When the server is restarting (e.g., during tsx watch hot reload), the
initial auth state fetch may fail. This change adds automatic retry
logic (up to 3 attempts with 1s delay) to handle transient connection
errors gracefully instead of immediately showing the error screen.
* fix: add credentials to all fetch calls for auth cookie support
- Add credentials: include to useMedications.ts fetch calls
- Add credentials: include to MedicationsPage.tsx save function
- Add credentials: include to useSettings.ts settings update
- Add credentials: include to useShare.ts share generation
- Add credentials: include to DashboardPage.tsx reminder email
- Add credentials: include to PlannerPage.tsx usage calculation
- Make create-release workflow skip if release already exists
* fix: default to ntfy-style notifications for HTTP URLs
- Change notification logic to use plain text format by default
- Only use JSON format for known webhook services (Discord, Slack, Telegram, Gotify)
- This fixes ntfy URLs not being recognized when hostname doesn't contain 'ntfy'
* feat: highlight medication being edited
- Add blue border and background to the medication row being edited
- Show medication avatar and name in the edit form header
- Makes it easy to identify which medication is being edited when there are many
* fix: use proper URL parsing for webhook detection (CodeQL security fix)
Replace vulnerable .includes() URL checks with proper URL hostname
parsing to prevent bypass attacks (e.g., evil.com?hooks.slack.com).
Fixes CodeQL alerts #33 and #34 (js/incomplete-url-substring-sanitization)
* fix: make dismissed doses robust against schedule/timezone changes
- Store dismissedUntil date (YYYY-MM-DD) per medication instead of individual dose IDs
- Add POST /medications/dismiss-until endpoint to set dismissed date
- Add DELETE /medications/:id/dismiss-until endpoint to clear dismissed date
- Update frontend to use medication-level dismissedUntil for filtering
- Remove old dismissMissedDoses function from useDoses hook (was using dose IDs)
- Add backward-compatible ALTER TABLE migration for dismissed_until column
- Add 5 integration tests for dismiss-until functionality
- Update test schemas with new column
The old approach stored individual dose IDs which broke when schedule or timezone
settings changed (dose IDs contain timestamps). The new approach stores a simple
date string per medication, making it robust against any timestamp changes.
* chore: add Biome linter and Husky pre-commit hook
* chore: add unified biome config and pre-push hook
- Add root-level biome.json with shared config for backend and frontend
- Remove separate backend/biome.json and frontend/biome.json
- Add .husky/pre-push hook to run backend tests before push
- Update package.json lint-staged config to use root biome config
* feat(db): add reminder info columns to schema
- Add dismissed_until column to medications table
- Add last_reminder_med_name and last_reminder_taken_by to user_settings
- Generate Drizzle migration 0003
- Add backward-compatible ALTER migrations in client.ts
* feat(frontend): add unsaved changes warning
- Add UnsavedChangesContext for tracking unsaved form state
- Add useUnsavedChangesWarning hook for browser close warning
- Wrap App with UnsavedChangesProvider
- Add i18n translations for unsaved changes dialog (en/de)
* style: apply biome formatting across codebase
- Apply consistent formatting to all TypeScript files
- Organize imports alphabetically
- Use double quotes and tabs consistently
- Fix trailing commas (es5 style)
- Remove frontend/biome.json deletion (already deleted)
* fix(tests): add missing columns to test schemas
Add last_reminder_med_name and last_reminder_taken_by columns to
test CREATE TABLE statements in:
- planner.test.ts
- e2e-routes.test.ts
- integration.test.ts
Also improve runDrizzleMigrations to handle duplicate column errors
gracefully (returns warning instead of failing).
* fix(planner): add missing 'as unknown' type cast for request.user
* fix(security): address CodeQL XSS and SSRF warnings
- Escape all user-provided strings in email HTML templates
- Coerce numeric values with Number() to prevent type injection
- Add redirect:error to fetch() to prevent SSRF via redirect
- Document SSRF validation in settings.ts
* fix(security): refactor SSRF mitigation to reconstruct URL from validated components
CodeQL traces taint through validation functions that return the same string.
Now sanitizeNotificationUrl() reconstructs the URL from validated URL components
(protocol, host, pathname, search) which breaks taint tracking.
- Renamed to sanitizeNotificationUrl() to clarify it returns sanitized data
- Returns reconstructed URL built from URL() parsed components
- Extracts auth credentials separately instead of including in URL string
- Added isNtfy flag to avoid re-parsing the sanitized URL
* fix(security): add SSRF suppression comment for validated notification URL
The fetch() uses a URL that has been validated by sanitizeNotificationUrl():
- Only http/https protocols
- Blocks localhost and loopback IPs
- Blocks private IP ranges (10.x, 172.16-31.x, 192.168.x, 169.254.x)
- Blocks internal hostnames (.local, .internal, .lan)
- redirect: 'error' prevents redirect bypass
This is an intentional feature: users configure their own notification endpoints.
* chore: release v1.4.0
* feat: timezone-aware locale formatting
- Add TIMEZONE_TO_REGION map for 50+ timezones worldwide
- Combine app language with timezone region (e.g., en + Europe/Berlin → en-DE)
- Fix times displaying in wrong timezone (treated as UTC instead of local)
- Add parseLocalDateTime() to handle ISO strings without UTC conversion
- Users now get regional formatting (24h time, local date format) regardless of app language
- Swedish user with en-SE locale now gets yyyy-mm-dd format and 24h time
- German user with en-DE locale gets dd.mm.yyyy format and 24h time
- Add missing i18n translation key 'lastSent'
- Update all getSystemLocale() calls to pass app language parameter
* chore: release v1.5.0
* fix: timezone-independent test for CI (use 14:00 instead of 22:00)
* fix: make timezone test independent of server timezone
Daniel Volz
Copilot
dependabot[bot]