The legacy ingest batch holds the per-work-order lock, taken in id order,
for every work order it may cancel until the batch commits. A vendor
uplift request now runs under the same lock. Uplift creation on both
routes refuses a work order cancelled by either lifecycle or status text,
so a request can neither slip past a cancel nor land after one.
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>
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.
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.
Read the duration from the MP4/MOV movie header (moov/mvhd) on both the
dispatcher media endpoint and the vendor portal completion upload, so a
direct request cannot bypass the browser check. Unreadable metadata still
never blocks an upload.
The 10-photo / 3-video limit only counted dispatcher attachments, so vendor
portal uploads could push a work order past it (and vice versa). Count the
work order's current vendor documents alongside its attachments on both the
dispatcher media endpoint and the vendor portal. A new completion version
does not count the version it replaces.
A misleading file name could move a file into a larger size class: a real
PDF or JPEG named .mp4/.mov got the 100 MB video cap on the vendor portal,
and a .jpg declared with a foreign video type got it on the media endpoint.
Classify by the same resolved type the signature check validates; the
extension only decides when no allowlisted type is known.
The Elastic Beanstalk nginx proxy kept its 1 MB default body limit, so every
media upload over ~1 MB got an nginx 413 before reaching the API. Ship a
.platform nginx override (120M) in the bundle and assert it in the bundle
contract.
Apply the client-confirmed contract: photos up to 10 MB (JPEG/PNG/HEIC),
videos up to 100 MB (MP4/MOV), at most 10 photos and 3 videos per work order,
with stable generic rejection messages. The request ceiling (110 MB) sits
between the per-kind caps and the proxy so oversize files get the generic
message. The vendor portal accepts the same photo/video types and caps.
* refactor(api): enforce service and data-service boundaries
* refactor(api): complete feature service boundaries
* refactor(identity): enforce service and data boundaries
* refactor(vendors): enforce service and data boundaries
* refactor(workorders): enforce service and data boundaries
* refactor(backend): enforce architecture and optimize dispatch
* style(backend): format changed architecture files
* fix(architecture): address backend review follow-ups
* fix(backend): sanitize exception disclosure in changed API endpoints
Replace raw exception-message disclosure (ex.Message) returned to API
callers with a stable sanitized public message plus correlated structured
internal logging, across the endpoints changed in this PR.
- Add SanitizedErrors helper: logs the original exception at Error with a
generated correlation id and returns a stable public message referencing
it so support can trace without exposing internals.
- Inject ILogger<T> into the 14 changed controllers and route every
ex.Message/dbex.Message disclosure through the helper, preserving status
codes, response shapes, and business data (e.g. OpenWorkOrders).
- Leave FluentValidation (vex.Errors) and existing fixed-message catches
untouched; out-of-scope controllers (Account/Contact/Employee/Asset/
PMSchedule) are unchanged.
- Add focused tests proving internal exception text is not returned and
that Error logging carrying the original exception is invoked.
* fix(architecture): abstract job run state access
* style: format board update service
* test: use collection assertion idiom
