- Forgot Password is limited to 3 codes an hour and 10 a day per email, and
an account gets 10 failed code checks a day across every code it is sent,
so new client addresses and new codes no longer buy more guesses. Refused
requests answer exactly like accepted ones.
- The reset email is queued to a background sender, and unregistered
addresses store a row no code can match, so both paths do the same work
and return without waiting on the mail provider. Each request also clears
expired codes.
- Code hashes are HMAC-SHA256 under a key derived with HKDF from the JWT
signing secret; rows in the previous unkeyed format stop matching.
- Email and code are read only from the JSON body.
Two concurrent first requests can leave two pending codes for one email.
Exhausting or using one now deletes all of them, so a sibling code cannot
become live afterwards.
Forgot Password answers every address the same way and emails a code only
to an active account. Codes are stored as salted SHA-256 hashes, expire 15
minutes after issue, are replaced by a newer request, and are checked only
against the email they were issued to. Five failed checks delete the code;
attempts are reserved with one conditional UPDATE so concurrent guesses
cannot exceed the budget. VerificationCode requires the email, and email and
code are accepted in the JSON body so they stay out of URLs.
The three anonymous endpoints are rate limited to 10 requests per 15
minutes per client IP. Forwarded headers are trusted only through loopback
and private hops, since the API sits behind the EB load balancer and nginx.
The migration adds hash, salt, expiry and attempt columns and deletes the
old plaintext rows.
User creation, contact, calendar event and site-contact create/modify/delete stamps used local server time. Validation rules comparing user-entered dates and the JWT expiry are unchanged.
GET api/team-members/me/permissions returns the keys the signed-in user
holds after role defaults and their own overrides, evaluated by the same
policy that guards writes. The user comes from the token; a missing or
unknown identity gets 401.
- Legacy reads (by id, all, by client, paged, address book, exists, count) and the vendor
preference site check now skip tombstoned sites
- Create requires a client, street address, city, state and at least one complete contact
UpdateAsync only applied WorkOrderType when it had a value, so once a
template was restricted to one work order type no request could make it
trade-generic again. The DTO now records whether workOrderType was present
in the body: omitting it keeps the stored value, an explicit null clears
it, and a concrete value sets it. The templates page echoes the stored
legacy fields on PUT, so its behaviour is unchanged.
- Reject duplicate site codes per client (case-insensitive); site code is immutable once set
- Delete tombstones the site and requires the DeleteSites permission (Admin, Scheduler)
- GET /api/locations/{id}/open-work-orders returns the open count and ids
- PATCH /api/locations/{id}/contact-info saves contacts and notes from the work-order Site dialog
- Add nullable Locations.Notes, used as the site-level POC notes fallback
Reactive/Emergency work orders with a SEV 1-5 level are timed from their
creation against the SEV Respond deadline (2/4/8/24/72 hours, one backend
table). From 50% they are at risk: a dismissable High row in the "SLA at
Risk" section and an entry in the feed's slaAtRisk set with the server
clock (start, deadline, percent) for the banner and toast. From 100% they
are a Critical acknowledge row that only acknowledging removes.
POST /api/notifications/sla/{id}/acknowledge records who and when as a
work-order audit entry ("SLA breach acknowledged by <name>"), scoped to the
caller's feed audience: 404 outside it, 409 before the deadline, 204 when
recorded or already recorded. A later severity change is a new breach.
Both hosts now apply the same Identity password rule: at least 6 characters
with one uppercase letter, one number and one special character. Change
password requires an authenticated caller, verifies the current password
before evaluating the new one, and reports a policy rejection separately
from a wrong current password.
Adds an extra safety note and an ordered procedure list to completion
document templates, name search, creator and last-updated audit fields,
a tenant-scoped count of open work orders that depend on a template, and
a delete that unlinks Services while they keep requiring a document.
Writes are gated by the create/edit/delete completion template team
permissions instead of the Admin role.
GET /board/search accepts ids=1,2,3 (positive ints, deduplicated, at most
200). When present the result is exactly those work orders inside the
caller's tenant and base scope; date, status, dispatcher, facet and text
filters are ignored so none of them can hide a listed work order.
Malformed or oversized lists are a 400.
The parity test now also asserts which rows the drill-down lists: legacy
open rows (status in Status or in LegacyStatus, or none) are listed and
legacy closed, cancelled or assigned rows are not. Drops ticket keys
from comments.
A vendor could withdraw (or cancel) an uplift a dispatcher raised from the work
order. Withdraw and its cancel alias now refuse requests with createdby set,
using the portal's not-found response, and the portal read model reports
RaisedByVendor so the portal can hide Revise and Withdraw on those requests.
A work-order request stores the requested increase, not a total. Letting
the vendor revise one after changes were requested rewrote RequestedNTE
as a total while it still read as a work-order request, corrupting its
amount and the NTE it would be approved to. Revise now answers not-found
for any request the vendor did not raise and leaves the row untouched.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Work-order requests store the requested increase in RequestedNTE; vendor
portal requests store the requested NTE total. The queue Delta, the
pending and approved exposure totals, the work-order uplift list, the
board summary and the notification Delta now all read one definition
(UpliftAmount) that honours both meanings and translates to SQL.
Approving a work-order request now adds its increase to the dispatch NTE
instead of replacing the NTE with the increase; vendor requests still end
at their requested total.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The approvals header summed the whole RequestedNTE for work-order-path
requests, while each Pending row shows RequestedNTE - CurrentNTE. When a
work order already had an NTE the header overstated exposure by that NTE.
The header now sums the same Delta the rows display, over the same rows
the Pending tab lists (non-deleted request on a non-deleted dispatch).
The unused duplicate aggregate is removed so one definition remains.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The strict date range made undated open work orders disappear for every
client that predates the includeDateless flag. The frontend on main sends
the all-weeks window (2000-01-01..2099-12-31) for "no range", the
1970-01-01..2099-12-31 window for the pinned Unassigned queue, and no date
input at all for the WO# duplicate lookup. None of those send the flag, so
deploying this backend before the frontend would have dropped undated WOs
from all three flows.
includeDateless is now optional. An explicit value still wins. When it is
omitted, a request with no date input or with an all-weeks Custom window
keeps its open undated rows, as before SH-391. Any other range stays
strict. The backend and frontend can therefore deploy in either order.
AddMedia sized a file before validating its type, and ValidateSize's Unknown
arm returned the video message, so a 150 MB .exe sent as
application/octet-stream was rejected as FileTooLarge with "Videos must be
100 MB or smaller." EnsureAllowed now runs first, so an unsupported file always
reports UnsupportedMediaType, and the Unknown arm uses a type-neutral message.
The oversize controller tests now use real file headers so they pass the type
check before reaching the size cap.
Approve, reject, request-changes, expiry and escalation staged their work
order audit with WorkOrderId = dispatch.WorkOrderId ?? 0. WorkOrderAuditLogs
requires a real work order, so any uplift on a dispatch without an owning
work order failed to save: the decision returned a 500 and the request stayed
Pending, and the expiry sweep failed on it every run.
The audit now goes to the work order the uplift resolves to through the
existing owner-or-linked read that revoke already uses. When none resolves,
the status change is saved on the request and no audit row is written.
The 10-photo / 3-video cap was a check-then-insert with no lock on both
upload surfaces, so two overlapping uploads could both take the last slot.
The board media upload and the vendor portal upload now run count, insert
and save inside ExecuteWorkOrderMutationAsync. GetMediaContent maps .heic
to image/heic.
UserCanApprove only accepted roles listed in Approvals:Tier1Roles/Tier2Roles,
so an environment whose config omits Admin denied every approval surface to
admins: can-approve, per-row CanDecide, approve/reject/request-changes and
evidence download. Admin now short-circuits the check; tier-role config still
governs every other role.
Adds MediaCounts (limits, current photos/videos, and the counts a new
completion version would see) so the portal can refuse an 11th photo or
4th video before uploading it. Uses the same count and replacement rule
the upload check enforces.
