Remove "value" from the market_value fallback chain in the SimpleFIN
HoldingsProcessor and add it to the cost_basis fallback chain instead.
Some brokerages (Vanguard, Fidelity) use "value" to represent cost basis,
causing the system to display average cost per share as the current price
and show massive phantom losses.
https://claude.ai/code/session_01V2gC6BPT3sF7Hu4XQgUQT4
Co-authored-by: Claude <noreply@anthropic.com>
* Reorganize import UI with Financial Tools / Raw Data tabs
Split the flat list of import sources into two tabbed sections using
DS::Tabs: "Financial Tools" (Mint, Quicken/QIF, YNAB coming soon) and
"Raw Data" (transactions, investments, accounts, categories, rules,
documents). This prepares for adding more tool-specific importers
without cluttering the list.
https://claude.ai/code/session_01BM4SBWNhATqoKTEvy3qTS3
* Fix import controller test to account for YNAB coming soon entry
The new YNAB "coming soon" disabled entry adds a 5th aria-disabled
element to the import dialog.
https://claude.ai/code/session_01BM4SBWNhATqoKTEvy3qTS3
* Fix system tests to click Raw Data tab before selecting import type
Transaction, trade, and account imports are now under the Raw Data tab
and need an explicit tab click before the buttons are visible.
https://claude.ai/code/session_01BM4SBWNhATqoKTEvy3qTS3
* feat: Add bulk import for NDJSON export files
Implements an import flow that accepts the full all.ndjson file from data exports,
allowing users to restore their complete data including:
- Accounts with accountable types
- Categories with parent relationships
- Tags and merchants
- Transactions with category, merchant, and tag references
- Trades with securities
- Valuations
- Budgets and budget categories
- Rules with conditions and actions (including compound conditions)
Key changes:
- Add BulkImport model extending Import base class
- Add Family::DataImporter to handle NDJSON parsing and import logic
- Update imports controller and views to support NDJSON workflow
- Skip configuration/mapping steps for structured NDJSON imports
- Add i18n translations for bulk import UI
- Add tests for BulkImport and DataImporter
* fix: Fix category import and test query issues
- Add default lucide_icon ("shapes") for categories when not provided
- Fix valuation test to use proper ActiveRecord joins syntax
* Linter errors
* fix: Add default color for tags when not provided in import
* fix: Add default kind for transactions when not provided in import
* Fix test
* Fix tests
* Fix remaining merge conflicts from PR 766 cherry-pick
Resolve conflict markers in test fixtures and clean up BulkImport
entry in new.html.erb to use the _import_option partial consistently.
https://claude.ai/code/session_01BM4SBWNhATqoKTEvy3qTS3
* Import Sure `.ndjson`
* Remove `.ndjson` import from raw data
* Fix support for Sure "bulk" import from old branch
* Linter
* Fix CI test
* Fix more CI tests
* Fix tests
* Fix tests / move PDF import to first tab
* Remove redundant title
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Add AI merchant enhancement and dedup
* Enhancements
Add error if job is already running
add note that we also merge merchants
* Allow updating provider website
* Review fixes
* Update provider_merchant.rb
* Linter and fixes
* FIX transaction quick menu modal
* Initial split transaction support
* Add support to unsplit and edit split
* Update show.html.erb
* FIX address reviews
* Improve UX
* Update show.html.erb
* Reviews
* Update edit.html.erb
* Add parent category to dialog
* Update en.yml
* Add UI indication to totals
* FIX ui update
* Add category select like rest of app
* Add split ui
* Add settings configuration for split transactions
- Adds a new settings section for appearance changes
- Also adds extra checks for delete and API calls
- Also adds checks for parent/child changes
* fixes
- split transactions dark mode fix
- add split transactions to context menu
* Update entry.rb
1. New validation split_child_date_matches_parent — prevents saving a split child with a date different from its parent. This is the root-cause fix that
protects all flows at once.
2. Bulk update guard — bulk_update! now strips :date from attributes when processing split children, preventing the validation from raising and silently
skipping the date change instead.
* N+1 fix for split_parent?
* Update entry.rb
Problem: In bulk_update!, when a split child has :date removed from attrs (line 432) and the remaining attrs is empty (e.g., the bulk update only
changed the date), entry.update! {} still ran as a no-op. But lock_saved_attributes! and mark_user_modified! at lines 443-444 executed unconditionally,
incorrectly marking untouched split children as user-modified and opting them out of future syncs.
Fix:
1. Added a changed flag to track whether any actual modification happened
2. Wrapped entry.update! in an if attrs.present? check so no-op updates are skipped
3. Gated lock_saved_attributes! and mark_user_modified! behind if changed, so they only run when the entry was actually modified (either via attribute
update or tag update)
* fixes
1. Indentation in show.html.erb Settings section — The split button block and delete block had extra indentation making them appear nested inside guard
blocks they weren't part of. Fixed to match actual nesting.
2. Skip @split_parents query when grouping is off — The controller now only loads split parent entries when show_split_grouped? is true, saving a query
with joins when the feature is disabled.
The "transaction merchant" condition filter used `family.assigned_merchants`
which only returned merchants already assigned to a transaction. This meant
newly created merchants wouldn't appear in the rule condition dropdown until
manually assigned to a transaction.
Changed to `family.available_merchants` which includes all family merchants
(even unassigned ones), provider merchants on transactions, and recently
unlinked merchants — consistent with the transaction form merchant selector.
Fixes#1197
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* Initial split transaction support
* Add support to unsplit and edit split
* Update show.html.erb
* FIX address reviews
* Improve UX
* Update show.html.erb
* Reviews
* Update edit.html.erb
* Add parent category to dialog
* Update en.yml
* Add UI indication to totals
* FIX ui update
* Add category select like rest of app
---------
Signed-off-by: Juan José Mata <juanjo.mata@gmail.com>
Co-authored-by: Juan José Mata <juanjo.mata@gmail.com>
* Add conditional migration for vector_store_chunks table
Creates the pgvector-backed chunks table when VECTOR_STORE_PROVIDER=pgvector.
Enables the vector extension, adds store_id/file_id indexes, and uses
vector(1024) column type for embeddings.
* Add VectorStore::Embeddable concern for text extraction and embedding
Shared concern providing extract_text (PDF via pdf-reader, plain-text as-is),
paragraph-boundary chunking (~2000 chars, ~200 overlap), and embed/embed_batch
via OpenAI-compatible /v1/embeddings endpoint using Faraday. Configurable via
EMBEDDING_MODEL, EMBEDDING_URI_BASE, with fallback to OPENAI_* env vars.
* Implement VectorStore::Pgvector adapter with raw SQL
Replaces the stub with a full implementation using
ActiveRecord::Base.connection with parameterized binds. Supports
create_store, delete_store, upload_file (extract+chunk+embed+insert),
remove_file, and cosine-similarity search via the <=> operator.
* Add registry test for pgvector adapter selection
* Configure pgvector in compose.example.ai.yml
Switch db image to pgvector/pgvector:pg16, add VECTOR_STORE_PROVIDER,
EMBEDDING_MODEL, and EMBEDDING_DIMENSIONS env vars, and include
nomic-embed-text in Ollama's pre-loaded models.
* Update pgvector docs from scaffolded to ready
Document env vars, embedding model setup, pgvector Docker image
requirement, and Ollama pull instructions.
* Address PR review feedback
- Migration: remove env guard, use pgvector_available? check so it runs
on plain Postgres (CI) but creates the table on pgvector-capable servers.
Add NOT NULL constraints on content/embedding/metadata, unique index on
(store_id, file_id, chunk_index).
- Pgvector adapter: wrap chunk inserts in a DB transaction to prevent
partial file writes. Override supported_extensions to match formats
that extract_text can actually parse.
- Embeddable: add hard_split fallback for paragraphs exceeding CHUNK_SIZE
to avoid overflowing embedding model token limits.
* Bump schema version to include vector_store_chunks migration
CI uses db:schema:load which checks the version — without this bump,
the migration is detected as pending and tests fail to start.
* Update 20260316120000_create_vector_store_chunks.rb
---------
Co-authored-by: sokiee <sokysrm@gmail.com>
DeveloperMessage was a debug-only STI subclass of Message that was never
created by any production code. Remove the model, view partial, test,
and fixtures, and simplify Chat#conversation_messages accordingly.
https://claude.ai/code/session_012pm5HKGKFs1tpAsvXMr4Tp
Co-authored-by: Claude <noreply@anthropic.com>
* feat(balance): incremental ForwardCalculator — only recalculate from changed date forward
When a Sync record carries a window_start_date, ForwardCalculator now
seeds its starting balances from the persisted DB balance for
window_start_date - 1, then iterates only from window_start_date to
calc_end_date. This avoids recomputing every daily balance on a
long-lived account when a single transaction changes.
Key changes:
- Account::Syncer passes sync.window_start_date to Balance::Materializer
- Balance::Materializer accepts window_start_date and forwards it to
ForwardCalculator; purge_stale_balances uses opening_anchor_date as the
lower bound in incremental mode so pre-window balances are not deleted
- Balance::ForwardCalculator accepts window_start_date; resolve_starting_balances
loads end_cash_balance/end_non_cash_balance from the prior DB record and
falls back to full recalculation when no prior record exists
- Tests added for incremental correctness, fallback behaviour, and purge safety
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
# Conflicts:
# app/models/balance/materializer.rb
* Enhance fallback logic on ForwardCalculator and Materializer
* fix(balance): address CodeRabbit review issues on incremental ForwardCalculator
- materializer.rb: handle empty sorted_balances in incremental mode by still
purging stale tail balances beyond window_start_date - 1, preventing orphaned
future rows when a transaction is deleted and the recalc window produces no rows
- materializer_test.rb: stub incremental? alongside calculate in the incremental
sync test so the guard in ForwardCalculator#incremental? doesn't raise when
@fell_back is nil (never set because calculate was stubbed out)
- materializer_test.rb: correct window_start_date in the fallback test from
3.days.ago to 2.days.ago so window_start_date - 1 hits a date with no
persisted balance, correctly triggering full recalculation instead of
accidentally seeding from the stale wrong_pre_window balance
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* feat(balance): multi-currency fallback to full recalculation and add corresponding tests
* address coderabbit comment about test
* Make the foreign-currency precondition explicit in the test setup.
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
* feat: Add default account for manual transaction entries (#1061)
Allow users to designate a default account that auto-selects
in the transaction creation form. Also consolidates account list
actions (edit, link/unlink, enable/disable) into a meatball menu.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* - handle context menu width on mobile
- restrict default account to depository types only
- added FR, ES and DE i18n files
* - Add credit card accounts can also be used as default
- Moved logic into controller
* Scope context menu max-width to accounts menu only
- decouples the width constraint from the shared DS::Menu component by introducing an optional max_width param
* fix ci test and address issues raised by coderabbit and codex
* Address CodeRabbit review feedback
- Use .present? for institution_name guards to avoid empty UI artifacts
- Align "Set default" menu visibility with actual preselection eligibility
(active + unlinked + supports_default?) to prevent drift between UI and model
- Keep disabled star visible when account is already default but now ineligible
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Add eligible_for_transaction_default? predicate to Account model
Consolidates active + unlinked + supports_default? checks into a single
shared predicate used by the controller, view, and user model guard,
preventing a direct PATCH from bypassing UI eligibility rules.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Added "Unset default" option
Added negative test for default account
Removed duplicated logic for account.eligible_for_transaction_default
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* Document admin-only reset auth in OpenAPI docs
The DELETE /api/v1/users/reset endpoint now requires admin role
(ensure_admin). Update the rswag spec to:
- Set default user role to admin so the 200 test passes
- Add a 403 response case for non-admin users with read_write scope
- Clarify the description notes admin requirement
- Add SuccessMessage schema and users paths to openapi.yaml
https://claude.ai/code/session_01Tj8ToLRmVg5HLmHwq9KKDY
* Consolidate duplicate 403 responses for reset endpoint
OpenAPI keys responses by status code, so two 403 blocks caused the
first (insufficient scope) to be silently overwritten by the second
(non-admin). Merge into a single 403 whose description covers both
causes: requires read_write scope and admin role. The test exercises
the read-only key path which hits 403 via scope check.
https://claude.ai/code/session_01Tj8ToLRmVg5HLmHwq9KKDY
* Em-dash out of messages.
* Fix tests
* Fix tests
---------
Co-authored-by: Claude <noreply@anthropic.com>
* feat(transaction): add support for file attachments using Active Storage
* feat(attachments): implement transaction attachments with upload, show, and delete functionality
* feat(attachments): enhance attachment upload functionality to support multiple files and improved error handling
* feat(attachments): add attachment upload form and display functionality in transaction views
* feat(attachments): implement attachment validation for count, size, and content type; enhance upload form with validation hints
* fix(attachments): use correct UI components
* feat(attachments): Implement Turbo Stream responses for creating and deleting transaction attachments.
* fix(attachments): include auth in activestorage controller
* test(attachments): add test coverage for turbostream and auth
* feat(attachments): extract strings to i18n
* fix(attachments): ensure only newly added attachments are purged when transaction validation fails.
* fix(attachments): validate attachment params
* refactor(attachments): use stimulus declarative actions
* fix(attachments): add auth for other representations
* refactor(attachments): use Browse component for attachment uploads
* fix(attachments): reject empty values on attachment upload
* fix(attachments): hide the upload form if reached max uploads
* fix(attachments): correctly purge only newly added attachments on upload failure
* fix(attachments): ensure attachment count limit is respected within a transaction lock
* fix(attachments): update attachment parameter handling to avoid `ParameterMissing` errors.
* fix(components): adjust icon_only logic for buttonish
---------
Signed-off-by: Juan José Mata <juanjo.mata@gmail.com>
Co-authored-by: Juan José Mata <juanjo.mata@gmail.com>
* refactor: use a map of providers that support pending transactions
* feat: add pending transaction manual merging tool
* fix(coderabbit): validate posted_entry_id against eligible posted candidates server-side
* fix(coderabbit): validate offset for negative numbers
* fix(coderabbit): check if pending_duplicate_candidates has_more in one transaction
* refactor: use list of radio buttons for better pagination
* chore: show current transaction range in paginated view
* chore: whitespace
chore: whitespace
* Feat: Add QIF (Quicken Interchange Format) import functionality
- Add the ability to import QIF files for users coming from Quicken
- Includes categories and tags
- Comprehensive tests for QifImport, including parsing, row generation, and import functionality.
- Ensure handling of hierarchical categories (ex "Home:Home Improvement" is imported as Parent:Child)
* Fix QIF import issues raised in code review
- Fix two-digit year windowing in QIF date parser (e.g. '99 → 1999, not 2099)
- Fix ArgumentError from invalid `undef: :raise` encoding option
- Nil-safe `leaf_category_name` with blank guard and `.to_s` coercion
- Memoize `qif_account_type` to avoid re-parsing the full QIF file
- Add strong parameters (`selection_params`) to QifCategorySelectionsController
- Wrap all mutations in DB transactions in uploads and category-selections controllers
- Skip unchanged tag rows (only write rows where tags actually differ)
- Replace hardcoded strings with i18n keys across QIF views and nav
- Fix potentially colliding checkbox/label IDs in category selection view
- Improve keyboard accessibility: use semantic `<label>` for file picker area
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix QIF import test count and Brakeman mass assignment warning
- Update ImportsControllerTest to expect 4 disabled import options (was 3),
accounting for the new QIF import type added in this branch
- Remove :account_id from upload_params permit list; it was never accessed
through strong params (always via params.dig with Current.family scope),
so this resolves the Brakeman high-confidence mass assignment warning
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix: QIF import security, safety, and i18n issues raised in code review
- Added french, spanish and german translations for newly added i18n keys
- Replace params.dig(:import, :account_id) with a proper strong-params
accessor (import_account_id) in UploadsController to satisfy Rails
parameter filtering requirements
- Guard ImportsController#show against QIF imports reaching the publish
screen before a file has been uploaded, preventing an unrescued error
on publish
- Gate the QIF "Clean" nav step link on import.uploaded? to prevent
routing to CleansController with an unconfigured import (which would
raise "Unknown import type: QifImport" via ImportsHelper)
- Replace hard-coded "txn" pluralize calls in the category/tag selection
view with t(".txn_count") and add pluralization keys to the locale file
- Localize all hard-coded strings in the QIF upload section of
uploads/show.html.erb and add corresponding en.yml keys
- Convert the CSV upload drop zone from a clickable <div> (JS-only) to
a semantic <label> element, making it keyboard-accessible without
JavaScript
* Fix: missing translations keys
* Add icon mapping and random color assignment to new categories
* fix a lint issue
* Add a warning about splits and some plumbing for future support.
Updated locales.
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
* Add post-trial inactive family cleanup with data archival
Families that expire their trial without subscribing now get cleaned up
daily. Empty families (no accounts) are destroyed immediately after a
14-day grace period. Families with meaningful data (12+ transactions,
some recent) get their data exported as NDJSON/ZIP to an ArchivedExport
record before deletion, downloadable via a token-based URL for 90 days.
- Add InactiveFamilyCleanerJob (scheduled daily at 4 AM, managed mode only)
- Add ArchivedExport model with token-based downloads
- Add inactive_trial_for_cleanup scope and requires_data_archive? to Family
- Extend DataCleanerJob to purge expired archived exports
- Add ArchivedExportsController for unauthenticated token downloads
https://claude.ai/code/session_01LR3Vo83R5s5SczYe6T33dQ
* Fix Brakeman redirect warning in ArchivedExportsController
Use rails_blob_path instead of redirecting directly to the ActiveStorage
attachment, which avoids the allow_other_host: true open redirect.
https://claude.ai/code/session_01LR3Vo83R5s5SczYe6T33dQ
* Update schema.rb with archived_exports table
Add the archived_exports table definition to schema.rb to match
the pending migration, unblocking CI tests.
https://claude.ai/code/session_01LR3Vo83R5s5SczYe6T33dQ
* Fix broken CI tests for ArchivedExports and InactiveFamilyCleaner
- ArchivedExportsController 404 test: use assert_response :not_found
instead of assert_raises since Rails rescues RecordNotFound in
integration tests and returns a 404 response.
- InactiveFamilyCleanerJob test: remove assert_no_difference on
Family.count since the inactive_trial fixture gets cleaned up by
the job. The test intent is to verify the active family survives,
which is checked by assert Family.exists?.
https://claude.ai/code/session_01LR3Vo83R5s5SczYe6T33dQ
* Wrap ArchivedExport creation in a transaction
Ensure the ArchivedExport record and its file attachment succeed
atomically. If the attach fails, the transaction rolls back so no
orphaned record is left without an export file.
https://claude.ai/code/session_01LR3Vo83R5s5SczYe6T33dQ
* Store only a digest of the download token for ArchivedExport
Replace plaintext download_token column with download_token_digest
(SHA-256 hex). The raw token is generated via SecureRandom on create,
exposed transiently via attr_reader for use in emails/logs, and only
its digest is persisted. Lookup uses find_by_download_token! which
digests the incoming token before querying.
https://claude.ai/code/session_01LR3Vo83R5s5SczYe6T33dQ
* Remove raw download token from cleanup job logs
Log a truncated digest prefix instead of the raw token, which is the
sole credential for the unauthenticated download endpoint.
https://claude.ai/code/session_01LR3Vo83R5s5SczYe6T33dQ
* Fix empty assert_no_difference block in cleaner job test
Wrap the perform_now call with both assertions so the
ArchivedExport.count check actually exercises the job.
https://claude.ai/code/session_01LR3Vo83R5s5SczYe6T33dQ
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Add invited users with delete button to admin users page
Shows pending invitations per family below active users in /admin/users/.
Each invitation row has a red Delete button aligned with the role column.
Alt/option-clicking any Delete button changes all invitation button labels
to "Delete All" and destroys all pending invitations for that family.
- Add admin routes: DELETE /admin/invitations/:id and DELETE /admin/families/:id/invitations
- Add Admin::InvitationsController with destroy and destroy_all actions
- Load pending invitations grouped by family in users controller index
- Render invitation rows in a dashed-border tbody below active user rows
- Add admin-invitation-delete Stimulus controller for alt-click behavior
- Add i18n strings for invitation UI and flash messages
https://claude.ai/code/session_01F8WaH5TmtdUWwhHnVoQ6Gm
* Fix destroy_all using params[:id] from member route
The member route /admin/families/:id/invitations sets params[:id],
not params[:family_id], so Family.find was always receiving nil.
https://claude.ai/code/session_01F8WaH5TmtdUWwhHnVoQ6Gm
* Fix translation key in destroy_all to match locale
t(".success_all") looked up a nonexistent key; the locale defines
admin.invitations.destroy_all.success, so t(".success") is correct.
https://claude.ai/code/session_01F8WaH5TmtdUWwhHnVoQ6Gm
* Scope bulk delete to pending invitations and allow re-inviting emails
- destroy_all now uses family.invitations.pending.destroy_all so accepted
and expired invitation history is preserved
- Replace blanket email uniqueness validation with a custom check scoped
to pending invitations only, so the same email can be invited again
after an invitation is deleted or expires
https://claude.ai/code/session_01F8WaH5TmtdUWwhHnVoQ6Gm
* Drop unconditional unique DB index on invitations(email, family_id)
The model-level uniqueness check was already scoped to pending
invitations, but the blanket unique index on (email, family_id)
still caused ActiveRecord::RecordNotUnique when re-inviting an
email that had any historical invitation record in the same family
(e.g. after an accepted invite or after an account deletion).
Replace it with no DB-level unique constraint — the
no_duplicate_pending_invitation_in_family model validation is the
sole enforcer and correctly scopes uniqueness to pending rows only.
https://claude.ai/code/session_01F8WaH5TmtdUWwhHnVoQ6Gm
* Replace blanket unique index with partial unique index on pending invitations
Instead of dropping the DB-level uniqueness constraint entirely, replace
the unconditional unique index on (email, family_id) with a partial unique
index scoped to WHERE accepted_at IS NULL. This enforces the invariant at
the DB layer (no two non-accepted invitations for the same email in a
family) while allowing re-invites once a prior invitation has been accepted.
https://claude.ai/code/session_01F8WaH5TmtdUWwhHnVoQ6Gm
* Fix migration version and make remove_index reversible
- Change Migration[8.0] to Migration[7.2] to match the rest of the codebase
- Pass column names to remove_index so Rails can reconstruct the old index on rollback
https://claude.ai/code/session_01F8WaH5TmtdUWwhHnVoQ6Gm
---------
Signed-off-by: Juan José Mata <juanjo.mata@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
* fix: Handle conditional loading of Plaid Link script
* fix: Plaid accounts not linking on first sync
* fix: Handle Plaid script loading edge cases
* fix: Use connection token for disconnect safety and retry failed script loads
* fix: Destroy Plaid Link handler on controller disconnect
* fix: Add timeout to Plaid CDN script loader to prevent deadlocks
* Make categories global
This solves us A LOT of cash flow and budgeting problems.
* Update schema.rb
* Update auto_categorizer.rb
* Update income_statement.rb
* FIX budget sub-categories
* FIX sub-categories and tests
* Add 2 step migration
* Add default family selection for invite-only onboarding mode
When onboarding is set to invite-only, admins can now choose a default
family that new users without an invitation are automatically placed into
as members, instead of creating a new family for each signup.
https://claude.ai/code/session_01U9KgikKjV6xbyBZ5wMYsYx
* Restrict invite codes and onboarding settings to super_admin only
The Invite Codes section on /settings/hosting was visible to any
authenticated user via the show action, leaking all family names/IDs
through the default-family dropdown. This tightens access:
- Hide the entire Invite Codes section in the view behind super_admin?
- Add before_action :ensure_super_admin to InviteCodesController for
all actions (index, create, destroy), replacing the inline admin? check
- Add ensure_super_admin_for_onboarding filter on hostings#update that
blocks non-super_admin users from changing onboarding_state or
invite_only_default_family_id
https://claude.ai/code/session_01U9KgikKjV6xbyBZ5wMYsYx
* Fix tests for super_admin-only invite codes and onboarding settings
- Hostings controller test: sign in as sure_support_staff (super_admin)
for the onboarding_state update test, since ensure_super_admin_for_onboarding
now requires super_admin role
- Invite codes tests: use super_admin fixture for the success case and
verify that a regular admin gets redirected instead of raising StandardError
https://claude.ai/code/session_01U9KgikKjV6xbyBZ5wMYsYx
* Fix system test to use super_admin for self-hosting settings
The invite codes section is now only visible to super_admin users,
so the system test needs to sign in as sure_support_staff to find
the onboarding_state select element.
https://claude.ai/code/session_01U9KgikKjV6xbyBZ5wMYsYx
* Skip invite code requirement when a default family is configured
When onboarding is invite-only but a default family is set, the
claim_invite_code before_action was blocking registration before
the create action could assign the user to the default family.
Now invite_code_required? returns false when
invite_only_default_family_id is present, allowing codeless
signups to land in the configured default family.
https://claude.ai/code/session_01U9KgikKjV6xbyBZ5wMYsYx
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Enforce one pending invitation per email across all families
Users can only belong to one family, so allowing the same email to have
pending invitations from multiple families leads to ambiguous behavior.
Add a `no_other_pending_invitation` validation on create to prevent this.
Accepted and expired invitations from other families are not blocked.
Fixes#1172https://claude.ai/code/session_016fGqgha18jP48dhznm6k4m
* Normalize email before validation and use case-insensitive lookup
When ActiveRecord encryption is not configured, the email column stores
raw values preserving original casing. The prior validation used a direct
equality match which would miss case variants (e.g. Case@Test.com vs
case@test.com), leaving a gap in the cross-family uniqueness guarantee.
Fix by:
1. Adding a normalize_email callback that downcases/strips email before
validation, so all new records store lowercase consistently.
2. Using LOWER() in the SQL query for non-encrypted deployments to catch
any pre-existing mixed-case records.
https://claude.ai/code/session_016fGqgha18jP48dhznm6k4m
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Initial plan
* Fix nil references in Recording failed LLM usage code paths
Co-authored-by: jjmata <187772+jjmata@users.noreply.github.com>
* Replace error&.message with rescue-guarded safe_error_message helper
error&.message only guards against nil; it still raises when the error
object's .message implementation itself throws (e.g. OpenAI errors that
call data on nil). Replace with a safe_error_message helper that wraps
error&.message in a rescue block, returning a descriptive fallback
string on secondary failures. Apply the helper in both record_usage_error
(usage_recorder.rb) and record_llm_usage (openai.rb), including the
regex branch of extract_http_status_code in both files.
Co-authored-by: jjmata <187772+jjmata@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: jjmata <187772+jjmata@users.noreply.github.com>
* Enhance logging in search_family_files.rb
Added logging for search parameters and results in SearchFamilyFiles.
Signed-off-by: Juan José Mata <juanjo.mata@gmail.com>
* Log level should be `debug` not `warn` here
* Unguarded `trace&.update` patterns
* API concernts from CodeRabbit
---------
Signed-off-by: Juan José Mata <juanjo.mata@gmail.com>
Both Uncategorized and Other Investments are synthetic categories with
id=nil. When expense_totals_by_category indexes by category.id, Other
Investments overwrites Uncategorized at the nil key, causing uncategorized
actual spending to always return 0.
Use category.name as fallback key (id || name) to differentiate the two
synthetic categories in all hash builders and lookup sites.
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* Add new Date field when creating a new Account
* Fix german translation
* Update app/controllers/concerns/accountable_resource.rb
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
Signed-off-by: Michel Roegl-Brunner <73236783+michelroegl-brunner@users.noreply.github.com>
* Add missing opening_balance:date to update_params
* Change label text
---------
Signed-off-by: Michel Roegl-Brunner <73236783+michelroegl-brunner@users.noreply.github.com>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
* Feat: Implement manual sync prices functionality and enhance holdings display
* Feat: Enhance sync prices functionality with error handling and update UI components
* Feat: Update sync prices error handling and enhance Spanish locale messages
* Fix: Address CodeRabbit review feedback
- Set fallback @provider_error when prices_updated == 0 so turbo stream
never fails silently without a visible error message
- Move attr_reader :provider_error to class header in Price::Importer
for conventional placement alongside other attribute declarations
- Precompute @last_price_updated in controller (show + sync_prices)
instead of running a DB query directly inside ERB templates
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix: Replace bare rescue with explicit exception handling in turbo stream view
Bare `rescue` silently swallows all exceptions, making debugging impossible.
Match the pattern already used in show.html.erb: rescue ActiveRecord::RecordInvalid
explicitly, then catch StandardError with logging (message + backtrace) before
falling back to the unknown label.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix: Update test assertion to expect actual provider error message
The stub returns "Yahoo Finance rate limit exceeded" as the provider error.
After the @provider_error fallback fix, the controller now correctly surfaces
the real provider error when present (using .presence || fallback), so the
flash[:alert] is the actual error string, not the generic fallback.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix: Assert scoped security_ids in sync_prices materializer test
Replace loose stub with constructor expectation to verify that
Balance::Materializer is instantiated with the single-security scope.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* Fix: Assert holding remap in remap_security test
Add assertion that @holding.security_id is updated to the target
security after remap, covering the core command outcome.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* Fix: CI test failure - Update disconnect external assistant test to use env overrides
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Use `# pipelock:ignore Credential in URL` on the specific false
positive line instead of excluding all of client.rb from scanning.
The rest of the file is now scanned normally.
The `subcategories` method queries `WHERE parent_id = category.id`, but
for the synthetic uncategorized budget category, `category.id` is nil.
This caused `WHERE parent_id IS NULL` to match ALL top-level categories,
making them appear as subcategories of uncategorized. This inflated
actual_spending and produced a large negative available_to_spend.
Add a nil guard on category.id to return an empty relation for synthetic
categories.
Fixes#819
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* Add budget rollover: copy from previous month
When navigating to an uninitialized budget month, show a prompt
offering to copy amounts from the most recent initialized budget.
Copies budgeted_spending, expected_income, and all matching category
allocations. Also fixes over-allocation warning showing on uninitialized
budgets.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* Redirect copy_previous to categories wizard for review
Matches the normal budget setup flow (edit → categories → show)
so users can review/tweak copied allocations before confirming.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* Address code review: eager-load categories, guard against overwrite
- Add .includes(:budget_categories) to most_recent_initialized_budget
to avoid N+1 when copy_from! iterates source categories
- Guard copy_previous action against overwriting already-initialized
budgets (prevents crafted POST from clobbering existing data)
- Add i18n key for already_initialized flash message
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* Add invariant guards to copy_from! for defensive safety
Validate that source budget belongs to the same family and precedes
the target budget before copying. Protects against misuse from
other callers beyond the controller.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* Fix button overflow on small screens in copy previous prompt
Stack buttons vertically on mobile, side-by-side on sm+ breakpoint.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* feat(helm): add Pipelock ConfigMap, scanning config, and consolidate compose
- Add ConfigMap template rendering DLP, response scanning, MCP input/tool
scanning, and forward proxy settings from values
- Mount ConfigMap as /etc/pipelock/pipelock.yaml volume in deployment
- Add checksum/config annotation for automatic pod restart on config change
- Gate HTTPS_PROXY/HTTP_PROXY env injection on forwardProxy.enabled (skip
in MCP-only mode)
- Use hasKey for all boolean values to prevent Helm default swallowing false
- Single source of truth for ports (forwardProxy.port/mcpProxy.port)
- Pipelock-specific imagePullSecrets with fallback to app secrets
- Merge standalone compose.example.pipelock.yml into compose.example.ai.yml
- Add pipelock.example.yaml for Docker Compose users
- Add exclude-paths to CI workflow for locale file false positives
* Add external assistant support (OpenAI-compatible SSE proxy)
Allow self-hosted instances to delegate chat to an external AI agent
via an OpenAI-compatible streaming endpoint. Configurable per-family
through Settings UI or ASSISTANT_TYPE env override.
- Assistant::External::Client: SSE streaming HTTP client (no new gems)
- Settings UI with type selector, env lock indicator, config status
- Helm chart and Docker Compose env var support
- 45 tests covering client, config, routing, controller, integration
* Add session key routing, email allowlist, and config plumbing
Route to the actual OpenClaw session via x-openclaw-session-key header
instead of creating isolated sessions. Gate external assistant access
behind an email allowlist (EXTERNAL_ASSISTANT_ALLOWED_EMAILS env var).
Plumb session_key and allowedEmails through Helm chart, compose, and
env template.
* Add HTTPS_PROXY support to External::Client for Pipelock integration
Net::HTTP does not auto-read HTTPS_PROXY/HTTP_PROXY env vars (unlike
Faraday). Explicitly resolve proxy from environment in build_http so
outbound traffic to the external assistant routes through Pipelock's
forward proxy when enabled. Respects NO_PROXY for internal hosts.
* Add UI fields for external assistant config (Setting-backed with env fallback)
Follow the same pattern as OpenAI settings: database-backed Setting
fields with env var defaults. Self-hosters can now configure the
external assistant URL, token, and agent ID from the browser
(Settings > Self-Hosting > AI Assistant) instead of requiring env vars.
Fields disable when the corresponding env var is set.
* Improve external assistant UI labels and add help text
Change placeholder to generic OpenAI-compatible URL pattern. Add help
text under each field explaining where the values come from: URL from
agent provider, token for authentication, agent ID for multi-agent
routing.
* Add external assistant docs and fix URL help text
Add External AI Assistant section to docs/hosting/ai.md covering setup
(UI and env vars), how it works, Pipelock security scanning, access
control, and Docker Compose example. Drop "chat completions" jargon
from URL help text.
* Harden external assistant: retry logic, disconnect UI, error handling, and test coverage
- Add retry with backoff for transient network errors (no retry after streaming starts)
- Add disconnect button with confirmation modal in self-hosting settings
- Narrow rescue scope with fallback logging for unexpected errors
- Safe cleanup of partial responses on stream interruption
- Gate ai_available? on family assistant_type instead of OR-ing all providers
- Truncate conversation history to last 20 messages
- Proxy-aware HTTP client with NO_PROXY support
- Sanitize protocol to use generic headers (X-Agent-Id, X-Session-Key)
- Full test coverage for streaming, retries, proxy routing, config, and disconnect
* Exclude external assistant client from Pipelock scan-diff
False positive: `@token` instance variable flagged as "Credential in URL".
Temporary workaround until Pipelock supports inline suppression.
* Address review feedback: NO_PROXY boundary fix, SSE done flag, design tokens
- Fix NO_PROXY matching to require domain boundary (exact match or .suffix),
case-insensitive. Prevents badexample.com matching example.com.
- Add done flag to SSE streaming so read_body stops after [DONE]
- Move MAX_CONVERSATION_MESSAGES to class level
- Use bg-success/bg-destructive design tokens for status indicators
- Add rationale comment for pipelock scan exclusion
- Update docs last-updated date
* Address second round of review feedback
- Allowlist email comparison is now case-insensitive and nil-safe
- Cap SSE buffer at 1 MB to prevent memory blowup from malformed streams
- Don't expose upstream HTTP response body in user-facing errors (log it instead)
- Fix frozen string warning on buffer initialization
- Fix "builtin" typo in docs (should be "built-in")
* Protect completed responses from cleanup, sanitize error messages
- Don't destroy a fully streamed assistant message if post-stream
metadata update fails (only cleanup partial responses)
- Log raw connection/HTTP errors internally, show generic messages
to users to avoid leaking network/proxy details
- Update test assertions for new error message wording
* Fix SSE content guard and NO_PROXY test correctness
Use nil check instead of present? for SSE delta content to preserve
whitespace-only chunks (newlines, spaces) that can occur in code output.
Fix NO_PROXY test to use HTTP_PROXY matching the http:// client URL so
the proxy resolution and NO_PROXY bypass logic are actually exercised.
* Forward proxy credentials to Net::HTTP
Pass proxy_uri.user and proxy_uri.password to Net::HTTP.new so
authenticated proxies (http://user:pass@host:port) work correctly.
Without this, credentials parsed from the proxy URL were silently
dropped. Nil values are safe as positional args when no creds exist.
* Update pipelock integration to v0.3.1 with full scanning config
Bump Helm image tag from 0.2.7 to 0.3.1. Add missing security
sections to both the Helm ConfigMap and compose example config:
mcp_tool_policy, mcp_session_binding, and tool_chain_detection.
These protect the /mcp endpoint against tool injection, session
hijacking, and multi-step exfiltration chains.
Add version and mode fields to config files. Enable include_defaults
for DLP and response scanning to merge user patterns with the 35
built-in patterns. Remove redundant --mode CLI flag from the Helm
deployment template since mode is now in the config file.
* Fix: use cookie/crumb auth in healthy? chart endpoint check
The health check was calling /v8/finance/chart/AAPL via the plain
unauthenticated client. Yahoo Finance requires cookie + crumb
authentication on the chart endpoint, so the health check would
fail even when credentials are valid. Updated healthy? to use
fetch_cookie_and_crumb + authenticated_client, consistent with
fetch_security_prices and fetch_chart_data.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix: add cookie/crumb auth to all /v8/finance/chart/ calls
fetch_security_prices and fetch_chart_data (used for exchange rates)
were calling the chart endpoint without cookie/crumb authentication,
inconsistent with healthy? and fetch_security_info. Added auth to both,
including the same retry-on-Unauthorized pattern already used in
fetch_security_info.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Update user-agent strings in yahoo_finance.rb
Updated user-agent strings to reflect current browser versions
Signed-off-by: Serge L <serge@souritech.ca>
* Fix: Add stale-crumb retry to healthy? and fetch_chart_data
Yahoo Finance returns 200 OK with {"chart":{"error":{"code":"Unauthorized"}}}
when a cached crumb expires server-side. Both healthy? and fetch_chart_data
now mirror the retry pattern already in fetch_security_prices: detect the
Unauthorized body, clear the crumb cache, fetch fresh credentials, and
retry the request once. Adds a test for the healthy? retry path.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Refactor: Extract fetch_authenticated_chart helper to DRY crumb retry logic
The cookie/crumb fetch + stale-crumb retry pattern was duplicated across
healthy?, fetch_security_prices, and fetch_chart_data. Extract it into a
single private fetch_authenticated_chart(symbol, params) helper that
centralizes the retry logic; all three call sites now delegate to it.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix: Catch JSON::ParserError in fetch_chart_data rescue clause
After moving JSON.parse inside fetch_authenticated_chart, a malformed
Yahoo response would throw JSON::ParserError through fetch_chart_data's
rescue Faraday::Error, breaking the inverse currency pair fallback.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix: Raise AuthenticationError if retry still returns Unauthorized
After refreshing the crumb and retrying, if Yahoo still returns an
Unauthorized error body the helper now raises AuthenticationError instead
of silently returning the error payload. This prevents callers from
misinterpreting a persistent auth failure as missing chart data.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Fix: Raise AuthenticationError after failed retry in fetch_security_info
Mirrors the same post-retry Unauthorized check added to fetch_authenticated_chart.
Without this, a persistent auth failure on the quoteSummary endpoint would
surface as a generic "No security info found" error instead of an AuthenticationError.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Signed-off-by: Serge L <serge@souritech.ca>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
* feat: Add tag badge in filter window
* fix: validate Tag color attribute as hex format and increase transparency mix in border color
* fix: use fallback for tag color
Replace direct trace.update calls with client trace upserts so OpenAI provider is compatible with langfuse-ruby 0.1.6 behavior. Add richer warning logs that include full exception details for trace creation, trace upserts, and generation logging failures. Add tests for client-based trace upserts and detailed error logging.
* Add MCP server endpoint for external AI assistants
Expose Sure's Assistant::Function tools via JSON-RPC 2.0 at POST /mcp,
enabling external AI clients (Claude, GPT, etc.) to query financial data
through the Model Context Protocol.
- Bearer token auth via MCP_API_TOKEN / MCP_USER_EMAIL env vars
- JSON-RPC 2.0 with proper id threading, notification handling (204)
- Transient session (sessions.build) to prevent impersonation leaks
- Centralize function_classes in Assistant module
- Docker Compose example with Pipelock forward proxy
- 18 integration tests with scoped env (ClimateControl)
* Update compose for full Pipelock MCP reverse proxy integration
Use Pipelock's --mcp-listen/--mcp-upstream flags (PR #127) to run
bidirectional MCP scanning in the same container as the forward proxy.
External AI clients connect to port 8889, Pipelock scans requests
(DLP, injection, tool policy) and responses (injection, tool poisoning)
before forwarding to Sure's /mcp endpoint.
This supersedes the standalone compose in PR #1050.
* Fix compose --preset→--mode, add port 3000 trust comment, notification test
Review fixes:
- pipelock run uses --mode not --preset (would prevent stack startup)
- Document port 3000 exposes /mcp directly (auth still required)
- Add version requirement note for Pipelock MCP listener support
- Add test: tools/call sent as notification does not execute
Balance sheet totals and accountable type summaries used a SQL JOIN on
exchange_rates matching only today's date, which returned NULL (defaulting
to 1:1) when no rate existed for that exact date. This caused foreign
currency accounts to show incorrect totals.
Changes:
- Refactor BalanceSheet::AccountTotals to batch-fetch exchange rates via
ExchangeRate.rates_for, with provider fallback, instead of a SQL join
- Refactor Accountable.balance_money to use the same batch approach
- Add ExchangeRate.rates_for helper for deduplicated rate lookups
- Fix net worth chart query to fall back to the nearest future rate when
no historical rate exists for a given date
- Add composite index on accounts (family_id, status, accountable_type)
- Reuse nearest cached exchange rate within a 5-day lookback window
before calling the provider, preventing redundant API calls on
weekends and holidays when providers return prior-day rates
https://claude.ai/code/session_01GyssBJxQqdWnuYofQRjUu8
Co-authored-by: Claude <noreply@anthropic.com>
* fix: Include investment_contribution in transfer? check and protect transfer entries from sync
Transfer transactions with kind "investment_contribution" were not recognized
as transfers by the UI, causing missing +/- indicators, "Transfer" labels,
and showing regular transaction forms instead of transfer details.
Also adds user_modified: true to entries created via TransferMatchesController
and SetAsTransferOrPayment rule action to protect them from provider sync
overwrites, matching the existing behavior in Transfer::Creator.
https://claude.ai/code/session_019BZ5Z1aqKSK3cRdR81P5Jg
* fix: Centralize transfer/budget kind constants for consistent investment_contribution handling
Define TRANSFER_KINDS and BUDGET_EXCLUDED_KINDS on Transaction to eliminate
hard-coded kind lists scattered across filters, rules, and analytics code.
investment_contribution is now consistently treated as a transfer in search
filters, rule conditions, and UI display (via TRANSFER_KINDS), while budget
analytics correctly continue treating it as an expense (via BUDGET_EXCLUDED_KINDS).
https://claude.ai/code/session_019BZ5Z1aqKSK3cRdR81P5Jg
* fix: Update tests for consistent investment_contribution as transfer kind
- search_test: loan_payment is now in TRANSFER_KINDS, so uncategorized
filter correctly excludes it (same as funds_movement/cc_payment)
- condition_test: investment_contribution is now a transfer kind, so it
matches the transfer filter rather than expense filter
https://claude.ai/code/session_019BZ5Z1aqKSK3cRdR81P5Jg
* fix: Eliminate SQL injection warnings in Transaction::Search
Replace string-interpolated SQL with parameterized queries:
- totals: use sanitize_sql_array with ? placeholders
- apply_category_filter: pass TRANSFER_KINDS as bind parameter
- apply_type_filter: use where(kind:)/where.not(kind:) and
parameterized IN (?) for compound OR conditions
- Remove unused transfer_kinds_sql helper
https://claude.ai/code/session_019BZ5Z1aqKSK3cRdR81P5Jg
---------
Co-authored-by: Claude <noreply@anthropic.com>
* fix: locale-dependent category duplication bug
* fix: use family locale for investment contributions category to prevent duplicates and handle legacy data
* Remove v* tag trigger from flutter-build to fix double-runs
publish.yml already calls flutter-build via workflow_call on v* tags,
so the direct push trigger was causing duplicate workflow runs.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* Refactor mobile release asset flow
* fix: category uniqueness and workflow issues
* fix: fix test issue
* fix: solve test issue
* fix: resolve legacy problem
* fix: solve lint test issue
* fix: revert unrelated changes
---------
Co-authored-by: Juan José Mata <juanjo.mata@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
