The attachment grid's <img> thumbnail pointed at the same
download_attachment.php URL as the full-size original, so previewing a
multi-MB photo attachment cost a full multi-MB download just to render a
small grid preview. loading="lazy" only deferred off-screen images; it
never reduced per-image transfer size.
Generate a resized JPEG thumbnail (longest side capped at 300px) via GD
at upload time, from the same metadata-stripped image stripImageMetadata()
already produces, reusing its decompression-bomb guard (~40MP decode cap).
Store the thumbnail's filename in a new nullable ticket_attachments.
thumbnail_filename column (migration 005); NULL means no thumbnail exists
(non-image, GD unavailable, or an attachment predating this change) and
callers fall back to the full-size original.
download_attachment.php serves the thumbnail when requested via
?thumb=1 and one exists, falling back to the original otherwise. The
attachments grid now requests thumb=1 for its <img> preview; the
lightbox link is unchanged and still opens the full-size original.
delete_attachment.php removes the thumbnail file alongside the original,
and cleanup_orphan_uploads.php's orphan lookup now also matches
thumbnail_filename so generated thumbnails aren't swept up as orphans.
Verified against real MariaDB + GD: a 1600x1200 test JPEG produced a
300x225 thumbnail at ~1.8KB vs. the 52KB original (~29x smaller);
confirmed the serving logic picks the thumbnail for image attachments
with one, falls back to the original for a non-image attachment even
when thumb=1 is requested, and that the updated orphan-cleanup lookup
matches both the original and thumbnail filename (and correctly finds
neither for an unrelated filename).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0117oBw2jN4kALYeS8HPq4zV
processBulkOperation()'s docblock claimed the transaction "ensures
atomicity - either all tickets are updated or none are," but that's
only true when $atomic = true is passed, and the only real caller
(api/bulk_operation.php) never passes it — the actual default is
best-effort: per-ticket failures are skipped and recorded, and every
other ticket in the batch still commits. Reworded the docblock to
describe the actual default behavior and when $atomic changes it.
The model already collected per-ticket failure reasons into
$result['errors'] (dashboard.js's bulkResultMessage() already reads
data.errors to render them), but api/bulk_operation.php's success
response dropped that field entirely, so admins only ever saw a bare
"N succeeded, M failed" count with no way to see which tickets failed
or why. Added 'errors' to the response when present.
Verified against real MariaDB: a bulk_status operation against a Closed
ticket (no transition defined) and an Open ticket (Open->Pending
defined) correctly processed 1/1, and the API response now includes
errors: ["Ticket ...: transition not allowed (Closed -> Pending)"].
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0117oBw2jN4kALYeS8HPq4zV
processBulkOperation() validated each ticket's status/priority transition
against a pre-transaction snapshot (ticketsById) instead of re-reading
inside the transaction, so two concurrent bulk operations touching the
same ticket could both pass validation against stale data and one
transition could silently clobber the other. Add lockTicketForUpdate(),
which re-fetches a ticket via SELECT ... FOR UPDATE, and use it wherever
the loop needs current status/priority, removing the three redundant
re-reads from the stale snapshot in the bulk_close/bulk_priority/
bulk_status branches.
Verified against real MariaDB with two concurrent OS processes: the
second process blocked ~1.2s on the first's held row lock, then correctly
observed the first's committed status and rejected an otherwise-stale-
data-permitted invalid transition. Also regression-tested normal
bulk_close/bulk_priority operation on real tickets.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0117oBw2jN4kALYeS8HPq4zV
status_transitions already has a DB-level UNIQUE KEY on
(from_status, to_status), so a genuine duplicate pair was never
actually possible to insert — but hitting that constraint raw
surfaced as an opaque "An internal error occurred" to the admin
instead of a clear message, since manage_workflows.php only validated
from_status !== to_status before attempting the insert/update.
Added an explicit existence check before insert/update in both the
POST and PUT handlers (excluding the row's own ID on update), so the
common case — an admin re-adding or renaming into a pair that already
exists — gets a specific 409 with the conflicting pair named, instead
of a generic 500. Also added ORDER BY transition_id to
WorkflowModel::getAllTransitions() as a defense-in-depth backstop:
since it collapses rows into a PHP array keyed by
[from_status][to_status] with no defined winner otherwise, if the DB
constraint were ever weakened or bypassed, this at least makes which
row wins deterministic (most recently created).
Verified against a real running server + real MariaDB: creating a
duplicate active pair, a duplicate inactive pair, and updating a
different row into an existing pair are all correctly rejected with
the friendly message; updating a row to keep its own existing pair
succeeds; and a genuinely different pair still creates normally.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Lhz7pGMaoTfL5sdYS5XiKv
The setValue()/is_required/select-options half of this issue was
already fixed incidentally by #47's new api/ticket_custom_fields.php
endpoint. The remaining gap: createDefinition()/updateDefinition()
never validated field_type against the six values the schema's
enum() actually allows (text/textarea/select/checkbox/date/number), so
a malformed type could be stored via the admin API and break whatever
UI renders it later.
Added an ALLOWED_FIELD_TYPES allowlist check at the top of both
methods, returning the same ['success' => false, 'error' => ...] shape
they already use for a DB failure — api/custom_fields.php already
propagates that shape correctly with no changes needed there. Verified
against real MariaDB: an invalid field_type is rejected on both create
and update, while a valid one still succeeds.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Lhz7pGMaoTfL5sdYS5XiKv
TicketModel::updateVisibility() only updated the tickets row — it
never touched ticket_watchers. A user watching a public ticket that's
later made confidential/internal, and who isn't creator/assignee/
admin/in the new visibility_groups, kept receiving Matrix
notifications (title + redacted activity preview) about a ticket
canUserAccessTicket() would now reject them from opening directly.
After a successful visibility update, re-evaluates every current
watcher against the new visibility rules via the same
canUserAccessTicket() check the rest of the app uses, and removes any
who no longer qualify. Verified against real MariaDB: tightening to
confidential correctly drops watchers with no standing access while
keeping an admin watcher; tightening to internal with a specific group
correctly keeps a watcher in that group and drops one who isn't.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Lhz7pGMaoTfL5sdYS5XiKv
BulkOperationsModel's bulk_close/bulk_status paths had zero references
to NotificationHelper — the exact same status transition (e.g.
Open->Closed) silently produced no Matrix/watcher notification when
performed via bulk actions, while the single-ticket edit page and
Bearer API both notify on every status change. Separately, their
audit_log entries used a bare ['status' => 'Closed', ...] shape
instead of the {'status': {'from': X, 'to': Y}} shape every other
status-change path uses, which broke two downstream consumers:
TicketView.php's timeline fell back to a generic "updated this
ticket" instead of "updated status", and notifications.php's
$details['status']['from'] on a string produced a broken "? -> ?"
notification title.
Fixed the audit_log shape for both operation types, and added a
notification queue collected during the per-ticket loop and flushed
only after a successful commit (so atomic-mode rollback correctly
sends zero notifications, matching how nothing else about a rolled-
back batch takes effect either). Also fixed an incidental bug found
while matching this to the single-ticket path: update_ticket.php's
notifyWatchers() call never passed the ticket's visibility, silently
defaulting to 'public' and always including the shared notify list
even for confidential/internal tickets — the exact leak #71 fixed
elsewhere in NotificationHelper itself, just never reaching this
call site.
Verified against real MariaDB with a real local webhook-capturing
server: bulk_close correctly fires sendStatusChangeNotification() +
notifyWatchers() with the right old/new status and a redacted title
for a confidential ticket; audit_log rows show the correct {from,to}
shape; and an atomic-mode rollback (one ticket's transition invalid)
sends zero notifications and leaves both tickets unchanged.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Lhz7pGMaoTfL5sdYS5XiKv
sendMentionNotification(), called from add_comment.php, had no
visibility check at all — unlike sendCommentNotification()/
notifyWatchers() which redact the comment preview for non-public
tickets. Mentioning a user with zero standing access to a confidential
ticket (not creator/assignee/admin, not in visibility_groups) sent
them a Matrix DM with the full ticket title AND comment text — worse
than #46 since it's delivered directly to an individual rather than
diluted into a shared list.
add_comment.php now filters mentioned users through
canUserAccessTicket() before resolving Matrix IDs, skipping the
notification entirely for anyone without access (one of the two
options the issue names as acceptable). getMentionedUsers() needed to
start selecting is_admin and groups alongside user_id/username/
display_name, since canUserAccessTicket() requires them. Verified
against real MariaDB: a user mentioned on a confidential ticket they
don't own/aren't assigned to is correctly denied, a user in the
matching visibility_groups for an internal ticket is correctly
allowed, and the same user is correctly denied on a different internal
ticket whose group they're not in.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015nCxwFFsy8ouMWzn56rPVP
getVisibilityFilter() (dashboard list/stats) matched via
FIND_IN_SET(?, REPLACE(t.visibility_groups, ' ', '')) — stripping
spaces from the column but not from the bound group name — while
canUserAccessTicket() (single-ticket access) did a plain trim with no
space-stripping at all. For a group name containing a space (e.g. "IT
Support"), a member could open an internal ticket directly by URL but
never see it in their dashboard list or stats counts.
Now strips spaces from the bound parameter too, matching the column-
side normalization, so both paths agree. Verified against real
MariaDB: a ticket visible via canUserAccessTicket() for a
space-containing group is now also matched by getVisibilityFilter()'s
SQL, a wrong-group user is denied by both, and the plain no-space case
is unaffected.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015nCxwFFsy8ouMWzn56rPVP
AttachmentModel::getAttachments() had no LIMIT/OFFSET, so a ticket
with hundreds of attachments loaded and rendered every one of them in
a single API response and DOM grid, unbounded.
Added optional limit/offset to getAttachments(), matching the pattern
already used by CommentModel::getCommentsByTicketId(). The GET handler
in upload_attachment.php now accepts limit/offset (default 40, capped
at 100) and returns total/has_more alongside the page of attachments.
ticket.js's loadAttachments()/renderAttachments() now fetch and append
pages, showing a "Load more attachments (N remaining)" control when
more are available. Verified against real MariaDB with 12 attachments
across 3 pages of 5: no duplicates or gaps across pages, and the
legacy unlimited call (getAttachments($ticketId) with no
limit/offset) still returns everything unchanged.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015nCxwFFsy8ouMWzn56rPVP
addDependency()'s "already exists" check only matched the exact
(ticket_id, depends_on_id, dependency_type) tuple. A user could add
"A blocks B" from ticket A's page, then separately add "B blocked_by
A" from ticket B's page — wouldCreateCycle() correctly found no cycle
(both normalize to the same precedence edge), so the insert was
allowed, creating two DB rows describing one real relationship (shown
twice on ticket B's page: once under Dependencies, once under
Dependents).
Added an inverse-relationship check before the insert: blocks/
blocked_by are inverses of each other, relates_to is its own inverse
(symmetric). duplicates has no defined inverse type in the schema, so
both directions remain independently insertable, which is correct —
"A duplicates B" and "B duplicates A" are distinct claims.
Verified against a local MariaDB instance: the exact repro from the
issue (A blocks B, then B blocked_by A) is now rejected, relates_to's
symmetric case is rejected in both directions, and duplicates in
either direction is unaffected.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MGDKHiU5RJdo3dqQUDow3X
deleteOldLogs() ran a single unbounded DELETE. created_at is indexed
so row selection itself is cheap, but on a large qualifying set (first
run after enabling/changing AUDIT_LOG_RETENTION_DAYS, or after the
cron silently missed runs) an unbounded single-statement DELETE holds
row locks for the full duration — risking contention with the frequent
concurrent INSERTs the audit log receives from live traffic. Now
deletes in batches of 1000 (parameterized), looping until nothing
qualifies.
Verified against a local MariaDB instance with a batch size of 10
forcing multiple loop iterations: deleted exactly the stale rows,
left recent rows untouched, correct total count returned.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MGDKHiU5RJdo3dqQUDow3X
toggleActive() flipped is_active without touching next_run_at. If a
schedule was disabled while next_run_at was still in the future, then
re-enabled after that date had passed, the next cron tick saw
next_run_at <= NOW() and fired immediately — surprising for an admin
expecting a re-enabled "daily" schedule to wait until its next natural
occurrence. Now recomputes next_run_at from the current time when
transitioning to active, matching what a fresh schedule creation would
produce; disabling is unchanged.
Verified against a local MariaDB instance: re-enabling a schedule
whose next_run_at was in 2020 recomputed it to tomorrow at the
scheduled time; disabling leaves next_run_at untouched.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MGDKHiU5RJdo3dqQUDow3X
syncUserFromAuthelia() did a plain check-then-insert with no
transaction, so two simultaneous first-visit requests for the same
brand-new user (e.g. two tabs opened right after SSO login) could
race: the second INSERT hits users.username's UNIQUE KEY, which
mysqli throws on (uncaught, PHP 8.1+ default report mode) rather than
returning false. Switched to INSERT ... ON DUPLICATE KEY UPDATE
followed by a re-fetch by username, so the losing request updates the
winner's row instead of throwing.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MGDKHiU5RJdo3dqQUDow3X
Zero callers anywhere in the codebase — superseded by claimForRun(),
which the cron script actually uses and which additionally guards
against the double-fire race between concurrent cron invocations that
this method lacked. Removing it so a future reuse doesn't silently
reintroduce that race.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MGDKHiU5RJdo3dqQUDow3X
Found while verifying #21 against the live schema: the model writes
'completed_with_errors' (21 chars) when a bulk operation finishes with
per-ticket failures, but bulk_operations.status was varchar(20), so the
write failed with "Data too long for column 'status'".
This was latent — bulk status changes previously forced every transition
through, so failed was always 0. Now that they honour the Workflow
Designer, a partially-skipped batch is a normal outcome and hits it.
- migrations/001 widens the column to varchar(32) (idempotent).
- The baseline is updated to match, for fresh installs.
- The bookkeeping UPDATE is wrapped in a try/catch: it runs after the
ticket changes are committed, so an instance deployed ahead of its
migrations must not turn a completed operation into an error response.
Verified against the live database with a disposable-ticket harness:
comment-required rejection changes nothing, undefined transitions are
refused per ticket with a reason, allowed transitions still work, mixed
batches apply the valid half, and an already-Closed ticket is a no-op.
Bulk status changes previously bypassed the workflow entirely — the model
carried an explicit "admin-only escape hatch" note — so bulk edit could
drive tickets through transitions the designer forbids and skip comments
the designer requires.
BulkOperationsModel now applies the same rules as the single-ticket path:
- Transitions absent from status_transitions are refused per ticket and
reported with a reason, instead of being forced through.
- requires_comment is checked up front across the whole selection, so a
batch is rejected before any ticket is mutated rather than half-applied.
- The reason is persisted as a comment on each ticket changed, matching
what a single-ticket close records.
- Tickets already in the target status are a no-op success, not a failure.
requires_admin needs no extra check: api/bulk_operation.php already gates
the endpoint on admin.
Client: both bulk modals now collect a reason, the close path gets a real
modal instead of a bare confirm, and per-ticket skip reasons surface in
the result toast instead of a bare failure count.
Foundation for extending the Bearer API beyond create-only:
- api_keys gains a scope column (read | read_write); baseline schema updated
and the column applied to the live DB. Existing keys default to
read_write so the hwmon create key keeps working.
- ApiKeyModel: createKey() takes a validated scope; validateKey() always
surfaces scope (defaults read_write); getAllKeys() is paginated
({keys,total,page,perPage}, key_hash stripped).
- ApiKeyAuth: expose getKeyContext() (scope/key_name/created_by/api_key_id)
and requireScope() (403 on insufficient scope); existing return values
unchanged.
- create_ticket_api.php: require read_write scope (a read key can't create).
- Admin /admin/api-keys: scope selector on the create form, a scope column,
and pagination (revoked keys were stacking up).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- create_ticket_api.php: remove the wrong CREATE TABLE stub that broke a
fresh DB; generate collision-safe ticket_ids so a genuine id collision
isn't misreported as a duplicate and a hw alert dropped; stop leaking
raw DB errors; correct a reopen comment that falsely claimed refreshed
sensor data
- manage_recurring.php: fix next-run so create/edit no longer skips the
current period (monthly day-of-month this month, daily today if time
not passed, correct ISO weekday, month-length clamp); only recompute
on schedule changes to avoid double-fire
- export_tickets.php, audit_log.php: neutralize CSV formula injection
- revoke_api_key.php, generate_api_key.php: correct HTTP status codes and
stop the catch clobbering specific 4xx codes
- health.php: stop leaking PHP version / extension names / paths to
unauthenticated callers
- watch_ticket.php: define $data before use
- manage_templates/recurring/custom_fields: add audit logging for CRUD;
add recurring_ticket + custom_field to the audit entity whitelist
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- CustomFieldModel: assign ?? fallbacks to variables before bind_param
(by-reference args cannot be ?? expressions; fatal on PHP 8.2, custom
fields were uncreatable/uneditable)
- RecurringTicketModel::create: fix swapped bind type for schedule_type
(enum bound as int coerced 'daily' to 0, breaking the cron)
- TicketModel/CommentModel: bind varchar ticket_id as string not int so
the unique index is usable and leading-zero IDs match; ticket_watchers
(int column) left as integer
- TicketModel::deleteTicket: delete from custom_field_values (real table)
not the nonexistent ticket_custom_fields
- TicketModel search: honor literal '0'; never emit AGAINST('*') on
all-special-char input (fall back to LIKE)
- TicketModel::updateTicket: disambiguate not-found vs no-op vs genuine
optimistic-lock conflict on zero affected rows
- WorkflowModel: do not cache transitions/statuses on DB failure (a
transient error no longer blocks all status changes for the TTL)
- DependencyModel: filter linked tickets by visibility (new optional user
context params) to stop confidential metadata leaking via dependencies
- BulkOperationsModel: validate status/priority/assignee before mutating
- AuditLogModel: gate getClientIP forwarded headers on trusted proxies;
add missing action/entity types so audit-log filters work
- WorkflowModel: add transitionRequiresComment() accessor for enforcement
- CommentModel: stop leaking raw DB errors to clients (log instead)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Medium:
- create_ticket_api.php: environment tags were parsed with explode('][') which
left brackets on the first/last tag so the whitelist never matched, dropping
the env tag from the dedup hash — a [production] and [staging] issue with
otherwise-identical components could collide onto one ticket. Use a
bracket-aware regex.
- CommentModel::getThreadedCommentsPaged only fetched DIRECT children of root
comments, so when pagination is active, nested replies at depth 2-3 vanished
from the thread. Expand replies level-by-level (bounded to depth 3).
- StatsModel::getTicketsByAssignee ignored the visibility filter the rest of the
stats apply, so a non-admin's "by assignee" widget counted (leaked) confidential
tickets. Thread the same filter through.
- watch_ticket.php GET path returned watch state / watcher names / count for any
ticket with no access check (the POST path checks it) — added canUserAccessTicket.
- dashboard.js kanban: every card rendered as P4 because the [class*="lt-p"]
selector never matched the lt-badge-p1 class and the fallback didn't strip "P".
Extract the digit directly.
Low:
- audit_log.php CSV: "Log ID" column was always blank ($log['log_id'] vs the real
audit_id column). Use audit_id.
- check_duplicates.php: the graceful-degradation try/catch only covered the throw
path; guard the false-return (non-exception mysqli) path too.
- notifications.php: owner-who-is-also-@mentioned got two notifications for one
comment; drop the duplicate comment row when a mention covers the same comment.
- dashboard.js hover preview rendered "PP1" (doubled prefix); strip the leading P.
- markdown.js: code/inline-code restore used string replace, so $&, $$, $`, $' in
user code were treated as replacement patterns; use a function replacer. Also
removed an unused loop var.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- SynapseHelper: memoize username->Matrix-ID lookups per-request (incl. negative
results) and add an overall time budget to resolveUsernames() plus a 2s connect
timeout, so notifying N watchers with a slow/unreachable Synapse can't stall the
request for N x 5s. (Chosen over async/queue per maintainer.)
- DependencyModel: fix cycle detection treating 'blocks' and 'blocked_by' as the
same edge direction. They are inverse relationships (single row each, no mirror
row), so the traversal now walks a unified precedence graph (blocks: ticket->
depends_on; blocked_by: depends_on->ticket) and wouldCreateCycle normalizes the
new edge's direction. Prevents both false-positive and missed cycles.
- CacheHelper: anchor prefix-delete to exact key boundaries (bare prefix or
prefix + '_' + md5) so delete('workflow') can't wipe a 'workflow_rules' cache.
- RateLimitMiddleware: hold an exclusive flock across the per-IP counter's
read-modify-write so concurrent requests can't both read N and write N+1
(undercounting past the limit). Fails open if the file can't be locked.
- dashboard.js: kanban status update now uses lt.api.post (per no-raw-fetch
convention) and reverts the card AND the optimistic column counts on failure
(the old raw-fetch catch left the card moved without reverting).
- BulkOperationsModel: document that bulk_status/bulk_close intentionally bypass
workflow transition validation (admin override, by design).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- create_ticket_api.php: validate status (against TICKET_STATUSES) and
priority (numeric 1-5). A non-numeric priority previously cast to 0 and
escalated the ticket below P1 on the dedup/update path.
- manage_workflows.php: reject empty/invalid from_status/to_status on POST
and PUT (must be valid ticket statuses) so the workflow table can't be
populated with bogus transitions.
- TicketModel::getAllTickets: COUNT(*) OVER() rides on returned rows, so a
page past the last row returned total/pages = 0. Fall back to a direct
COUNT when an over-range page yields no rows, keeping pager math correct.
- DashboardView: stop double-escaping category/type/assigned active-filter
labels (they were htmlspecialchars'd into the label and again at output,
rendering R&D as R&D); output escaping is retained.
- check_duplicates.php / NotificationHelper::notifyWatchers: wrap the DB
lookups in try/catch so a failed prepare/query degrades gracefully
(advisory dup-check returns none; best-effort watcher notify is skipped)
instead of fataling the request. Works whether mysqli throws or returns
false. (manage_* endpoints already have a top-level try/catch.)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
XSS / security:
- markdown.js: sanitize footnote labels to a safe slug before using them in
id/href attributes. Labels are captured before the HTML-escape pass, so a
label like x"><img onerror=...> broke out → stored XSS (the earlier quote-
escape fix didn't cover this path). Verified neutralized.
- RateLimitMiddleware: only trust X-Forwarded-For / X-Real-IP when REMOTE_ADDR
is a configured trusted proxy, and use the rightmost (proxy-appended) entry.
Previously any client could rotate XFF to escape the per-IP rate limit.
- .env.example: document TRUSTED_PROXIES so fresh deploys aren't fail-open on
the Authelia forward-auth spoofing protection.
Correctness:
- notifications.php: my previous assigned-to LIKE fix anchored only on '}', so
BULK assignments (logged {"assigned_to":N,"bulk_operation_id":..}) produced
no "assigned to you" notification — now matches both '}' and ',' delimiters.
- notifications.php: implement the documented @mention notifications (query
action_type='mention' rows for the current user); they were never delivered.
- NotificationHelper::notifyWatchers: guard unchecked prepare() so a missing
ticket_watchers table can't fatal the request after its DB write committed.
- AuditLogModel::getTicketTimeline: JSON_UNQUOTE the extracted ticket_id so
comment events actually match (string vs JSON-number comparison never did).
- AuditLogModel/audit_log.php: CSV export no longer silently truncates to the
1000-row UI cap; uses a dedicated higher export limit.
- DashboardView: quick-preview drawer read .ticket-link from the title cell
(which has none), so the title was always blank — use the cell text.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Continued fixes from the multi-agent review:
- recurring tickets cron: now that the parse error is fixed the job runs,
exposing two latent bugs. (1) next_run_at was only advanced after the
full success path, so any failure (e.g. a NULL created_by passed to the
non-nullable assignTicket() $assignedBy -> TypeError) left it in the past
and re-created a duplicate ticket every cron cycle. Added an atomic
claimForRun() (conditional UPDATE gated on still-due) called BEFORE
creation, which also prevents overlapping runs from double-creating.
(2) The cron used a raw mysqli with no utf8mb4, corrupting non-ASCII
content; it now uses Database::getConnection(). Also guard the assignment
so created_by NULL falls back to the assignee.
- bulk delete: attachment files were unlinked inside the DB transaction, so
an atomic-mode rollback restored rows but the files were already gone.
deleteTicket() can now defer file removal to the caller, and
BulkOperationsModel deletes files only after a successful commit.
- UserModel: back-tick the `groups` column (reserved word on MySQL 8.0.2+).
- create_ticket_api.php: stop leaking raw DB/exception messages to callers;
log server-side and return a generic error. (Also includes a pre-existing
working-tree tweak that adds title to the manual-ticket dedupe hash.)
- CI: semgrep install failed under PEP 668; add --break-system-packages.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
PHP 8.2 strict mysqli mode throws mysqli_sql_exception on duplicate key
rather than returning false from execute(). Replace the old if/else errno
check with try/catch on mysqli_sql_exception, re-throw non-1062 errors,
and use random_int range 100000000-999999999 (no leading zeros) for the
retry ID.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
PHP 8.2 raises mysqli_sql_exception on prepare() for non-existent tables
rather than returning false. Wrap each child-table delete in try/catch and
silently skip tables that don't exist in all deployments, re-throwing for
unexpected errors.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
All ticket_id parameters were bound as integer ("i"), which stripped
leading zeros before insertion into ticket_attachments.ticket_id
(VARCHAR 9). This caused a mismatch: upload_attachment.php creates
the directory using the full string (e.g. /uploads/000123456/) but
the DB stored the integer form ("123456"), so download and delete
would look in the wrong path.
Changed getAttachments, addAttachment, getTotalSizeForTicket, and
getAttachmentCount to use string binding ("s") so the canonical
zero-padded ticket ID is stored and read back consistently.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- TicketModel: add deleteTicket() that removes all child records
(comments, watchers, dependencies, attachments, custom fields)
then deletes the ticket and cleans up physical attachment files
- BulkOperationsModel: add bulk_delete case to processBulkOperation()
so the "Bulk Delete" UI button actually works instead of silently
failing with N failures
- bulk_operation.php: validate operation_type against whitelist to
reject unknown operations early with a proper error
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
bulk_operation.php: ticket ID validation was converting IDs to int then back
to string, so '000123456' became '123456' which never matched the DB VARCHAR
key, silently rejecting ~11% of tickets from bulk operations. Now validates
with ctype_digit() to preserve leading zeros.
TicketModel::getTicketsByIds(): changed intval() to strval() and bind type
'i' to 's' so VARCHAR ticket_id columns are queried consistently as strings.
DashboardController::getCategoriesAndTypes(): added null check on query
result before calling fetch_assoc() to prevent TypeError if query fails.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Created Today tile: no longer limits to open statuses (count is all statuses)
- Closed Today tile: filters by closed_at range, not updated_at
- Add closed_from/closed_to support to TicketModel and DashboardController
- Add Created/Updated/Closed date range inputs to sidebar filter panel
- Apply button collects date inputs; Clear All removes them
- removeFilter handles date chip removal (clears both _from and _to)
- Active filter chips shown for date ranges
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- add_comment.php: include user_id in response for avatar rendering
- ticket.js: add buildCommentElement() helper that matches server-rendered
comment structure (avatar, edit/delete buttons, textarea); use it in
addComment() and submitReply() so new comments show the avatar immediately
- AuditLogModel: logCommentCreate uses action_type='comment' not 'create'
- TicketView: formatAction handles entity_type='comment' with action_type='create'
for existing DB records; prevents "created this ticket" showing for comments
- update_ticket.php: remove owner/assignee restriction so any authenticated
team member can update ticket status and fields
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- manage_recurring.php calculateNextRun(): expand monthly cap from 28→31
with proper last-day-of-month clamping (matches model fix); use split
with ':00' append to handle malformed time strings without crashing;
fix weekly day array to start at index 1 (not 0) so day=0 never maps
to empty string and blows up DateTime
- RecurringTicketModel::calculateNextRunTime(): same weekly day array fix
(start at index 1) to eliminate '' → DateTime exception on day=0
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- TicketView.php: Show 3 lt-skeleton-card placeholders in the comment list
while "Load more" fetches; skeletons are removed on resolve or error
- ticket.css: Add .comment-skeleton margin spacing
- WorkflowDesignerView.php + manage_workflows.php: Prevent creating/editing
status transitions where from_status === to_status (client + server check)
- RecurringTicketsView.php: Expand monthly day picker from 28 to 31 days
(days 29-31 labelled "last day in short months")
- RecurringTicketModel.php: Clamp monthly schedule day to last day of target
month using format('t') instead of hard-capping at 28
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
User input containing MySQL boolean operators (+, -, (, ), ~, *, ", @)
was passed directly to MATCH...AGAINST in BOOLEAN MODE, causing MySQL to
parse them as search operators rather than literals. Input like '(test)'
or '-keyword' would result in a MySQL syntax error / empty results.
Strip boolean mode special chars before building the FULLTEXT term;
the raw search string is still used unchanged for the LIKE fallback parts.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
next_run_at was typed 'i' (int) but stores a datetime string → should be 's'.
is_active was typed 's' (string) but stores 0/1 boolean → should be 'i'.
Positions 10-11 were swapped: 'ssssiiisssis' → 'ssssiiisssii'.
The UPDATE method already had the correct types; only INSERT was affected.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Dashboard stat cards now show lt-dot trend indicators (up/warn/idle) based on
created_today vs closed_today flow — no extra DB query needed
- Add collapsible Team Workload panel showing assignee open ticket counts with
progress bars (green/cyan/red by load), avatar, and name
- StatsModel.getTicketsByAssignee() now returns proper objects with user_id,
display_name, open_count (was name-keyed flat array); limit raised to 8
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Ticket watchers:
- api/watch_ticket.php: GET (watch state) + POST (watch/unwatch toggle)
- index.php: route for /api/watch_ticket.php
- TicketView: WATCH/UNWATCH button with live state fetch and toggle
- NotificationHelper::notifyWatchers(): fetches watchers from DB, resolves
Matrix IDs via Synapse, fires notification to watchers + global list
- add_comment.php, update_ticket.php: call notifyWatchers on comment and
status-change events respectively
Fulltext search:
- TicketModel::hasFulltextIndex(): detects FULLTEXT index via information_schema
- getAllTickets(): uses MATCH...AGAINST when fulltext index exists, LIKE fallback
when not yet applied — zero-downtime rollout
Single-query pagination:
- getAllTickets() replaces separate COUNT + SELECT with COUNT(*) OVER() window
function — one round trip to DB per page load instead of two
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Comment pagination:
- CommentModel: add getCommentCount(), paginated getCommentsByTicketId()
with getThreadedCommentsPaged() for threading + LIMIT/OFFSET
- TicketController: load first 50 root comments + total count on page load
- api/get_comments.php: new AJAX endpoint for Load More (index.php routed)
- TicketView: Load More button + buildCommentEl() JS renderer for AJAX comments;
passes totalComments/commentOffset/isAdmin to window.ticketData
Matrix integration:
- NotificationHelper: add sendStatusChangeNotification(), sendCommentNotification(),
sendMentionNotification(), sendAssignmentNotification() alongside existing
sendTicketNotification(); internal fire() helper replaces duplicated cURL logic
- SynapseHelper: new helper that resolves SSO usernames → Matrix IDs by querying
Synapse Admin REST API directly (no caching, no stale data)
- config.php: add SYNAPSE_ADMIN_URL, SYNAPSE_ADMIN_TOKEN, MATRIX_NOTIFY_COMMENTS,
MATRIX_NOTIFY_ASSIGNMENTS config keys (all from .env)
- api/update_ticket.php: fire status-change notification after successful save
- api/add_comment.php: resolve @mentioned usernames via SynapseHelper and fire
mention notification; fire general comment notification when MATRIX_NOTIFY_COMMENTS=1
- api/assign_ticket.php: fire assignment notification (resolves assignee via Synapse)
when MATRIX_NOTIFY_ASSIGNMENTS=1
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Bug fixes:
- bulk-delete action called undefined bulkDelete() — wired to the
existing showBulkDeleteModal() so the confirmation modal actually shows
UX:
- Template loader now checks for existing title/description and asks
for confirmation before overwriting user-typed content
- Visibility select shows a dynamic hint paragraph that updates when
the user changes the selection (public/internal/confidential)
Architecture:
- TICKET_STATUSES added to config as single source of truth; all
hardcoded ['Open','Pending','In Progress','Closed'] arrays in
DashboardView now read from config; bulk-status modal in dashboard.js
reads window.TICKET_STATUSES (set from PHP) with array fallback
- ASSET_VERSION now auto-computed from max mtime of dashboard/ticket
CSS+JS files so browsers always pick up changes on deploy; manual
override still available via ASSET_VERSION in .env
- Removed 10 dead standalone stat methods from StatsModel (getOpenTicketCount,
getClosedTicketCount, getTicketsByPriority, etc.) — all superseded by
the consolidated fetchAllStats() queries, never called externally
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- TicketModel::getAllTickets() now accepts optional $user param and applies
getVisibilityFilter() so non-admin users cannot see internal/confidential
tickets they lack access to from the dashboard listing
- DashboardController passes $GLOBALS['currentUser'] to getAllTickets()
- clone_ticket.php: move Content-Type header to top so all error paths send
correct JSON content type
- AuthMiddleware: filter group names from HTTP header to [a-z0-9_-] only,
preventing header injection via malformed group names
- add_comment.php: return HTTP 201 on success, 500 in catch block
- update_comment.php, delete_comment.php: return 500 in catch blocks
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- upload_attachment.php: derive stored file extension from validated MIME type
instead of user-supplied filename, preventing executable extension attacks
(e.g. a PHP file renamed to evil.txt would now be stored as .txt)
- CustomFieldModel.php: fix bind_param type string in updateDefinition()
'sssssiiiii' (10 chars) → 'sssssiiii' (9 chars) to match 9 SQL placeholders
- RateLimitMiddleware.php: replace MD5 with SHA256 for rate limit file hashing
- user_preferences.php: add httponly, secure, samesite=Lax flags to ticketsPerPage
cookie to prevent XSS/CSRF cookie theft
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
StatsModel queries used `FROM tickets WHERE` with no table alias, but
getVisibilityFilter() returns SQL referencing `t.visibility`. Admins
were unaffected because they get `1=1` with no column references.
Added `t` alias to all three tickets queries that use $visSQL.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- bulk_operation.php: replace is_numeric() with strict int cast+equality to reject scientific notation
- AttachmentModel.php: fix bind_param type strings (s→i for integer ticket IDs)
- CommentModel.php: use strict !== comparison with (int) cast for user_id ownership checks
- ticket.js: replace all non-TDS class names (text-amber→lt-text-amber, btn→lt-btn variants, etc.)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
StatsModel.getAllStats() now accepts a user array and applies the same
getVisibilityFilter() logic used by ticket listings. Admins continue to
share a single cached result; non-admin users get per-user cache entries
so confidential ticket counts are not leaked in dashboard stats.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
jaredandClaude Sonnet 5