v2 replaces pnpm with bun and moves the standalone build to
apps/nextjs/.output/standalone, which already carries static assets,
public and next.config.ts. Next still writes an incomplete
.next/standalone, so the assembler now refuses to run without the v2
output rather than silently shipping that one.
The notes now list v2's registration points, the tests that already
fail on the upstream tag, and the one-way v1 to v2 database migration.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
v2 makes httpAuth mandatory on every integration definition: it is the
server-owned authentication the Stats widget and arbitrary requests reuse.
Both APIs take the key in an X-API-KEY header; Site Manager reads it from
its own secret kind.
Also split the inline type imports v2's stricter oxlint rejects.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Le fixture déclarait warnings: [], inféré never[], alors que le test lui
assigne un tableau de chaînes. tsc refuse ; vitest ne le voit pas.
Ajouté à la vérification de chaque tâche : pnpm turbo typecheck, pas
seulement pnpm test.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Constat de revue : un échec total des appels de statut produit le même
tableau vide qu'un compte sans configuration. Le rendu reste honnête —
le tiret n'affirme rien — mais le commentaire, lui, tranchait à tort.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Scan pré-vol. Le test de garde du bundle client lisait son source via
__dirname, qui n'existe pas en ESM et n'est utilisé par aucun autre spec
du dépôt. Remplacé par import.meta.url, avec la directive d'environnement
node puisqu'il lit le disque et que le défaut est jsdom.
Ajouté aux contraintes globales : vitest n'inclut que **/*.spec.ts, donc
un fichier .spec.tsx ne serait jamais exécuté et son absence passerait
pour un succès.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
L'état des tunnels entre dans quatre des sept tâches : schémas et index
par site (tâche 2), deux méthodes de client bornées (tâche 3),
orchestration en enrichissement non fatal (tâche 4), colonne et
traductions (tâche 6), et une étape de vérification dédiée (tâche 7).
Deux points de justesse portés par des tests :
- `ping: 0` devient `pingMs: null`. Relevé sur un tunnel connecté entre
deux sites distants, où une latence nulle est physiquement
impossible : le champ veut dire « non mesuré ».
- le nombre de configurations interrogées est borné à cinq. Le statut
étant indexé par configuration, un compte à dix produirait onze
requêtes par rafraîchissement.
La topologie reste hors périmètre : une liste de quinze réseaux n'a pas
sa place dans une ligne de tableau.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Ajout demandé après validation : l'état des tunnels, indexé par siteId,
se greffe sur les lignes existantes. La topologie reste hors v1 — une
liste de quinze réseaux n'a pas sa place dans une ligne de tableau.
Relevé au passage : ping vaut 0 sur un tunnel connecté entre deux sites
distants, où une latence nulle est physiquement impossible. Le champ
signifie « non mesuré » et le contrat le traduit en null, plutôt que de
propager un zéro que l'interface afficherait comme une performance.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Le validateur de palette ne couvre que les palettes catégorielles. Sain,
dégradé et coupé sont des statuts : palette réservée et fixe, assortie de
la règle qu'ils ne voyagent jamais sans icône ni libellé.
Une icône par segment étant impossible sur vingt-quatre segments, la
bande est doublée d'un résumé textuel. Mesuré : un vert/orange/rouge
classique donne un ΔE de 2,9 entre orange et vert en vision protanope.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
L'utilisateur revient sur la forme liste et retient la proposition
initiale : tableau HomarrDataTable avec la bande de santé en colonne.
La bande reste dans la v1 (D4 inchangée) — c'est la forme qui change,
pas le périmètre. Le tableau rend le tri et la visibilité responsive,
absorbe la croissance du nombre de sites, et correspond à ce que fait
l'interface Ubiquiti.
Contrepartie consignée : dans une colonne la bande est étroite, ce qui
fait de l'infobulle par segment le moyen principal de lire une tranche
plutôt qu'un confort.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Proposition de l'utilisateur, retenue : fusionner les métriques ISP avec
la barre « Internet Health » pour une liste d'état rapide des sites.
Elle inverse la décision D4. La barre avait été écartée au motif que
`internetIssues` est un tableau creux dont il faudrait reconstituer les
périodes saines — vrai de cette source, mais `isp-metrics` fournit la
même information déjà dense, une entrée par site, vérifié. La barre
coûte alors une boucle de coloration, pas un moteur de graphe.
D3 change aussi : liste d'état plutôt que tableau, la bande au centre
plutôt qu'en colonne. HomarrDataTable n'a donc plus lieu d'être, et le
tri disparaît. Limite assumée et consignée : au-delà de cinq ou six
sites la forme tableau redeviendrait préférable.
Point de justesse ajouté au contrat : une tranche horaire sans mesure
vaut "unknown" et jamais "healthy". Un site enregistré depuis douze
heures n'a que douze tranches sur vingt-quatre ; les peindre en vert
affirmerait un bon fonctionnement jamais mesuré.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Nouvelle intégration cloud-only, distincte de unifiOs, remontant une vue
multi-sites. Premier des trois projets identifiés à partir des captures
du Site Manager ; il construit aussi le socle des deux suivants.
Constat notable : il n'existe pas d'API cachée. /ea/* est un alias
complet de /v1/*, et une trentaine d'autres chemins plausibles
renvoient 404. La surface documentée est toute la surface.
La séparation types / parse est imposée dès la conception par le défaut
rencontré sur l'intégration précédente, où un import serveur dans un
module réexporté côté client empêchait l'application de démarrer.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Trois demandes utilisateur, plus une adaptation rendue nécessaire par la
première.
- Ordre inversé : descendant avant montant.
- Rafraîchissement porté de 60 s à 5 s. Il fallait le faire sur les DEUX
couches locales : l'intervalle client seul aurait relu la même entrée
de cache serveur douze fois. Le cache cloud reste à 5 min, ses
métriques n'évoluant que par tranches de 5 min.
- Le graphe est rendu dès la première mesure : vide plutôt qu'absent,
ce qui évite un saut de mise en page et n'affirme rien de faux.
Adaptation : la console ne rafraîchit ses statistiques que toutes les
30 s (mesuré — `lastHeartbeatAt` avance par pas de 30 s exacts, valeurs
identiques entre deux battements). Interroger toutes les 5 s renvoie
donc six fois la même mesure. Sans filtrage, le graphe deviendrait un
escalier et 60 points ne couvriraient plus que 5 minutes. Une lecture
identique à la précédente n'est donc pas enregistrée.
Seuil de péremption locale recalibré de 150 s à 30 s : l'ancien était
dimensionné pour 60 s + 60 s et ne signalait plus rien d'utile.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
L'API UniFi ne fournit que l'instantané et Homarr ne stocke aucun
historique : la série est accumulée côté client, comme le fait déjà le
widget systemResources. Elle repart donc vide à chaque rechargement.
Deux décisions d'honnêteté, dans la ligne du reste du widget :
- une lecture en échec est ignorée, jamais enregistrée comme un zéro,
qui dessinerait une chute à la ligne de base n'ayant pas eu lieu ;
- sous deux points, on affiche les chiffres seuls : une ligne plate
tracée sur un point unique se lit comme un réseau au repos alors
qu'on n'a rien mesuré.
Couleurs validées par le validateur de palette (bande de luminosité,
chroma, séparation daltonisme et vision normale, contraste) en thème
clair ET sombre. La paire orange.5/yellow.5 du widget systemResources
n'a délibérément pas été reprise : son ΔE en vision normale est de
12,1, sous le plancher de lisibilité de 15.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The widget stub (packages/widgets/src/unifi-os/) was incorrectly added in
Task 6 and belongs to Task 7. Its removal confirms the type system does not
require it, and prevents inconsistencies: the stub declares IconWifi while
Task 7 specifies IconRouter, and it is not registered in manifest.ts,
catalog.ts or refetch-intervals.ts.
- Delete packages/widgets/src/unifi-os/ directory entirely
- Revert packages/widgets/src/index.tsx to commit 57286d83d
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Create unifiOs request handler with 60s cache and stale-on-error fallback, and tRPC router exposing clientApi.widget.unifiOs.summary query. Include minimal widget stub for type system.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Introduces UnifiOsIntegration, orchestrating UnifiOsApi (local) and
UnifiSmApi (optional cloud) to build the UnifiOsSummary. Local endpoint
calls degrade independently via Promise.allSettled; the cloud client is
never instantiated without a configured siteManagerApiKey secret.
Registers the unifiOs kind in the integration creator, index, and types
barrels, resolving the previously-failing typecheck.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Task 4 of the UniFi OS integration: an optional cloud client for
Ubiquiti's Site Manager API that enriches the local console with ISP
latency data. Instantiated only when a cloud API key is configured
(Task 5). Every failure path (bad key, non-ok status, parse error,
unmatched gateway MAC) returns null rather than throwing, since the
cloud is a pure enrichment. Metrics are cached 5 minutes, keyed by
gateway MAC, to respect the ~27 min publication lag and avoid rate
limiting; the query window is 3 hours because a 15-minute window
returns an empty array on the real API.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Implements HTTP client for UniFi OS console integration API with support for
fetching site information, client counts, device statistics, and VPN server
configs. Includes pagination handling with filter-based fallback when client
count exceeds API limit.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Guarded environment validation with SKIP_ENV_VALIDATION and SECRET_ENCRYPTION_KEY
set in vi.hoisted() so that tests pass both with 'pnpm test' (NODE_ENV=development)
and bare 'pnpm vitest run' (no env override).
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Plan en 9 tâches avec cycle TDD par étape, du vocabulaire dans
@homarr/definitions jusqu'au déploiement sur le LXC.
Corrige aussi la spec sur deux points relevés en écrivant le plan :
- catalog.ts (widgetCatalogIcons) est un cinquième point d'enregistrement,
vérifié par manifest.spec.ts et absent du §5.2
- le §9 décrivait un test d'agrégation multi-site devenu sans objet depuis
la décision de ne pas agréger
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Conception validée pour une intégration UniFi OS en complément de
unifiController, basée sur l'API d'intégration officielle (clé X-API-KEY)
plutôt que sur l'ancienne API du Network Controller.
Toutes les capacités et limites de l'API ont été mesurées sur une console
réelle en 10.6.102, pas déduites de la documentation publiée (qui s'arrête
à 10.4.57 et est déjà en retard sur les consoles).
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* feat: migrate tooling and runtimes to bun
* fix: serialize act installs and adapt backup checks
* fix: finish bun scripts and rootless act setup
* fix: support bun images in runtime benchmarks
* fix: keep a single bun toolchain configuration
* fix: include docs mascot sources in workshop build
* docs: use bun in remaining tool commands
* fix: complete bun sqlite development tooling
* docs: explain sqlite development adapter
* fix: align remaining commands and local ci validation
* fix: use renovate regex fallback with bun
* fix: finalize bun sqlite connections during cleanup
* fix: preserve sqlite foreign key enforcement
* fix: align docker build frontends
* fix: prepare release versions with bun
* Add Bun benchmarks and preserve SQLite connection defaults
* Benchmark Bun installs, conflicts and production workloads
* Force Bun workspace tools and HTTP/1.1 upstream streaming
* Exclude benchmark evidence from Docker contexts
* Expose branch benchmarks through CI
* Verify hydrated widget data in production benchmarks
* Measure image archives and document benchmark protocol
* Audit dependency sharing across every workspace
* Pin pnpm for nested baseline benchmark scripts
* Preserve nested manager pinning on benchmark resume
* Include application and runtime image footprints
* Fail benchmark gates on candidate command errors
* Report complete build overhead and sampled memory bounds
* Validate migration comparisons across pinned revisions
* Record configurable runtime memory budgets
* Retain out-of-memory kills in runtime budget probes
* Reject OOM events consistently across runtime metrics
* Record complete Bun migration benchmarks
* Record memory diagnosis and correct Turbo caching
* Use Bun package management while preserving Node runtimes
* Keep native build fallbacks and exclude development artifacts
* Fix Node ESM metadata and isolate local CI configuration
* Measure development performance and isolate act environment
* Support isolated environment files in rootless act
* Validate the full migration range with act
* Preserve native SQLite binding for bundled migrations
* Use documented dev defaults and report benchmark scope
* Validate dev defaults and isolate restart measurements
* Record package-manager and development benchmarks
* Keep benchmark measurements and link full command artifacts
* Clean up temporary act environment files
* Validate Node runtimes in package-manager comparisons
* Validate every workspace in manual CI runs
* Retain patched nested brace expansion versions
* Complete package-manager and Node performance report
* Normalize archived benchmark logs
* Record final hosted validation
* Cache both Next builds with platform isolation
* Prime cache with write-only mode
* Split Next artifacts below Cloudflare upload limit
* Assemble coherent standalone cache outputs
* Preserve traced middleware entrypoints
* Assemble traced directory links before files
* Separate cached dependencies from assembled output
* Record signed Turbo cache benchmarks
* Remove archived benchmark captures
Full benchmark evidence is preserved at 8f7d3bcd14109bfb13fa664d31710b88d820ad66.
* Remove migration benchmark harness
* fix: correct migration launch and local Act prerequisites
* Consolidate Turbo caches and production artifacts
* Exclude generated files from Docker cache inputs
* fix: address Bun migration review findings
* fix: check restored compiler bundle manifests
* feat(api): expand REST coverage for full dashboard automation
Closes#6435.
Boards could only be created empty through the API. The single endpoint for
adding a widget took no geometry, so the size came from the widget default and
the position was always the next free cell. Everything else a setup needs was
reachable from the interface only.
Board composition
- POST /api/boards/items accepts an optional placement: width, height, xOffset,
yOffset and sectionId, or a layouts array for per breakpoint control. Without
it the previous automatic placement is unchanged.
- Adds read, update and delete for items, CRUD for sections, read and replace
for layouts, and GET /api/boards/{id} for the whole board.
- GET /api/boards/{id}/export and POST /api/boards/import make a board
reproducible from version control. onConflict decides what happens when the
name is taken, so a pipeline can be re-run.
- Board access is manageable through GET/PUT /api/boards/{id}/permissions.
Remaining services
- Search engines, integrations, groups, api keys and server settings are
exposed over OpenAPI and registered as MCP tools.
- GET /api/config/export and POST /api/config/import cover the configuration of
a whole instance. Ids are kept so references inside opaque widget options stay
intact, and entities are matched by natural key when the id is unknown, which
is what makes a document from another instance applicable.
- The backup endpoints accept an ApiKey header, so a recovery job no longer
needs a browser session.
Headless bootstrap
- homarr api-keys create prints a key without opening the web interface, which
together with the existing recreate-admin removes the last manual step.
Notes
- No database migration, the schema already stores position and size.
- Grid coordinates and sizes in a document are bounded, and the placement scan
no longer walks every column, so a large grid cannot occupy the server.
- Imports collect their statements first and apply them in one transaction, so a
rejected document never leaves a board half written.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: harden dashboard automation APIs
* fix: validate global home board settings
* fix: validate merged import settings
* docs(api): describe the bounded placement scan accurately
The comment still described the earlier version that walked every row up to the
last occupied one. Both axes now iterate the edges of the occupied areas, so the
cost follows the amount of content rather than the dimensions of the grid.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix(api): keep the search engine type narrow and record the new MCP tools
The `.output()` added to `searchEngine.getPaginated` widened `type` back to a
string through the plain select schema, which broke the management page that
switches on it. It now reuses the narrowed union the detail endpoint already has.
`mcp.spec` pins the reviewed tool inventory; the 33 tools this branch exposes are
added to it. The configuration round trip asserted the first layout of an item,
which is the mobile one half of the time and holds the same item clamped, so it
addresses the base layout by id instead.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* feat(api): complete Homarr automation endpoints
* refactor(api): consolidate preferences and task docs
* refactor(api): narrow REST to dashboard automation
* refactor(api): expose existing dashboard controls only
* fix(docs): drop removed API key CLI path
---------
Co-authored-by: rfdrew <20803599+rfdrew@users.noreply.github.com>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Co-authored-by: Thomas Camlong <thomas@ajnart.fr>
Co-authored-by: ajnart <thomas@ajnart.dev>
* fix: use native DNS resolution by default for Docker services
* test: verify Docker service DNS through production Compose
* test: isolate Docker DNS smoke resources per invocation [skip ci]
Add two regression tests that fail on v1.76.0 and pass with the fix:
- wire-level test: a Headers instance (the shape @immich/sdk hands over)
must reach the remote server with all headers intact
- real-SDK e2e: getServerStatistics through fetchWithTrustedCertificatesAsync
against a live local server must deliver x-api-key
The v1.76.0 User-Agent injection object-spread `fetchOptions.headers`, which
silently drops all headers when the value is a Headers instance. The Immich
SDK passes a Headers object with the merged x-api-key header, so requests
lost their API key and returned 401 (issue #6672). Build a real Headers
instance instead, keeping case-insensitive user-agent handling.
Fix Unraid logical CPU averaging, preserve native context menus for app and bookmark widgets in view mode, and prevent scrollbar-width GridStack resize oscillation.
OMV reports filesystem devices (sda1) without the /dev/ prefix, so the
partition-suffix normalization never applied and filesystem-to-SMART
matching returned no temperature. Strip partition suffixes from
unprefixed sd/vd, nvme and mmcblk names, and use the shared normalized
matcher in both the health monitoring and system disks widgets.
* chore: migrate cursor rules to portable agent skills and add mantine MCP guidance
Replace Cursor-only .cursor/rules/*.mdc with Agent Skills open-standard
SKILL.md files in .agents/skills/ (canonical), mirrored to .cursor/skills/
and .claude/skills/ so they work across Claude Code, Cursor, Codex, and
OpenCode. Add mantine MCP server guidance to AGENTS.md and un-ignore
.agents/skills/ in .gitignore.
* refactor: single-source agent skills at .agents/skills
Cursor, Codex, and OpenCode all read .agents/skills/ natively, so the
.cursor/skills/ mirror was redundant. Claude Code reads .claude/skills/,
which is now a symlink back to .agents/skills/ so the content lives once.
Deletes .cursor/skills/ and converts .claude/skills/ to a symlink.
* chore: ship mantine MCP server via project-scoped .mcp.json
Declares the keyless @mantine/mcp-server in .mcp.json (Claude Code
project-scope standard, committable) and documents MCP server setup and
the mantine entry in AGENTS.md.
* chore: address coderabbit security-review and markdown lint findings
Add explicit security-review guidance before exposing MCP tools and label
plain-text code fences to satisfy markdownlint MD040.
* feat(http): set custom User-Agent on outbound integration requests
Set a default 'User-Agent: Homarr/<version> (+https://homarr.dev)' header
on all outbound HTTP requests made via fetchWithTrustedCertificatesAsync,
so Homarr's integration traffic is distinguishable in reverse-proxy,
WAF, and security-monitoring logs instead of appearing as generic undici.
Closes#6604
* refactor(version): centralize Homarr version in @homarr/version package
Replace fragile deep-relative 'package.json' imports across the monorepo
with a single source of truth: the new @homarr/version leaf package reads
the version once from the root package.json and exports it as VERSION.
Consumers now import by package name incl. the new User-Agent header in
fetchWithTrustedCertificatesAsync (packages/core):
- core, api (info router), analytics, request-handler (update checker)
This also removes all manually-maintained relative version imports.
* fix(http): normalize User-Agent injection with Headers
Object-spreading fetchOptions.headers breaks when callers pass a Headers
instance (spread yields no entries) or a tuple array (numeric keys), and is
case-sensitive. Build a Headers instance and set the default User-Agent only
when no case-insensitive User-Agent entry exists, via a shared helper in both
fetch paths.
Adds unit tests covering record, Headers, tuple-array, and casing variants.
* revert(version): undo version refactor and tests, keep single-file change
Keep the User-Agent change as a single edit to
packages/core/src/infrastructure/http/request.ts. Remove the @homarr/version
package, the consumer migrations, and the unit tests.
* fix: keep nginx IPv6 listen only when IPv6 is available (#4596)
Commit 2c48b1896 (1.45.2) made the internal nginx proxy listen on the IPv6
wildcard ('listen [::]:7575;'), which aborts nginx startup on hosts/VMs with
IPv6 disabled:
nginx: [emerg] socket() [::]:7575 failed (97: Address family not supported by protocol)
A plain revert would reintroduce #2697 and break IPv6-only/dual-stack
environments where this listener is required. Instead, only emit the IPv6
listen directive when the host actually has IPv6 configured.
Reuse the envsubst step run.sh already performs: compute NGINX_LISTEN_IPV6
from /proc/net/if_inet6 and substitute it into the template. proc files always
report size 0, so presence is detected by grepping the file (not 'test -s').
* fix: honor HOMARR_DISABLE_IPV6 override in nginx config generation
Review feedback on #6539: give users an explicit opt-out from the IPv6
listener even when the host reports IPv6. The issue reporter (#4596)
requested such an env var, and the pattern matches the existing
ENABLE_DNS_CACHING IPv6 toggle. Auto-detection behavior is unchanged when
the var is unset. Validated in containers: override wins, dual-stack
preserved by default, IPv4-only fallback on IPv6-less hosts.
* ci: drop --delete-branch from reconcile auto-merge call
The reconcile job runs without a git checkout, so gh pr merge --delete-branch
fails with 'not a git repository' before enabling auto-merge. Auto-merge
itself needs no local git state; branch cleanup is already handled by the
PR-creating workflows.
* ci: checkout repo in bot auto-merge workflows
gh pr merge fails with 'not a git repository' when no checkout is present,
which prevented the reconcile job from enabling auto-merge. Restore the
checkout the previous approval workflow relied on.
The reconcile job runs without a git checkout, so gh pr merge --delete-branch
fails with 'not a git repository' before enabling auto-merge. Auto-merge
itself needs no local git state; branch cleanup is already handled by the
PR-creating workflows.
* ci: auto-approve and auto-merge trusted bot pull requests
Replace the approve-only automatic-approval workflow with a combined
approve + enable-auto-merge workflow that also re-runs on every push,
plus a scheduled reconcile job that re-enables auto-merge for prepared
bot PRs if GitHub ever disables it. Bot PRs are gated to the known bot
actors, same-repo head, and base branch dev.
* ci: match bot authors by app slug in reconcile filter
gh pr list reports pull request authors of GitHub Apps as app/<slug>;
adjust the allowlist so the reconcile job actually matches bot PRs.
Remove incompatible persisted widget query data before resetting a crashed widget so it can fetch fresh data instead of re-rendering the same stale payload.
* fix(immich): use thumbnail endpoint instead of original for album carousel
Switches from /api/assets/:id/original to /api/assets/:id/thumbnail to fix
HEIC image incompatibility (thumbnails are always JPEG), reduce bandwidth,
and improve loading speed.
Closes#6388
* working on immich
* Improve image proxy error handling
* fix(image-proxy): use trusted fetch response type
* docs(immich): require Asset View permission
The pnpm fetch docs recommend `pnpm install -r --offline` without
--frozen-lockfile. The --offline flag already enforces no registry
access, making --frozen-lockfile redundant.
https://pnpm.io/cli/fetch
PR #6296 introduced pnpm fetch in the Dockerfile for Docker cache
optimization. However, pnpm fetch crashes with ENOENT when
patchedDependencies are configured with nodeLinker: hoisted because
applyPatchToDir tries to chdir into node_modules/<pkg> which doesn't
exist during fetch (only the content-addressable store is populated).
Fix: temporarily switch nodeLinker to isolated during pnpm fetch so
patches apply inside the virtual store (node_modules/.pnpm/...) which
IS created by pnpm fetch. Restore hoisted before the install step so
the final node_modules layout stays flat.
Refs: pnpm/pnpm#5268, pnpm/pnpm#8726
* fix: context menu refresh stays open with loading spinner, cache age reacts to refetch
* style: auto-fix lint and format
* refactor: pass isFetching as prop to WidgetCacheAge to avoid duplicate useIsFetching
---------
Co-authored-by: ajnart <ajnart@users.noreply.github.com>
Next.js 16.2.9 does not fully support TypeScript 7 (tsgo) as the
primary typescript package. The Go-native compiler has no in-process
JavaScript API — lib/typescript.js is missing — which causes Next.js's
hasNecessaryDependencies check to fail and the build to SIGABRT with
CI=true. Vercel is working on native tsgo integration expected in Q3
2026 alongside TypeScript 7 stable.
The Mantine Indicator component (used by the Proxmox cluster health
widget for VM/LXC status dots) sits at z-index 200, which is higher
than the AppShellHeader's default 100. This caused the status dots
to bleed above the header when scrolling.
Set the header z-index to 201 so it stacks above Indicator content
without conflicting with modal/popover layers (300+).
Closes#6194
* fix(widget-secrets): add missing widget_secret migration and add error notifications
The widget_secret table was defined in all three Drizzle schema files but
no SQL migration was ever generated to create it, causing setSecret to
fail with 'no such table: widget_secret' in production (drizzle-kit push
in development hid the problem).
- Add migration SQL for sqlite (0040), mysql (0042), postgresql (0008)
- Update journal entries for all three dialects
- Add onError handlers to setSecret/deleteSecret mutations so errors
are shown as red notifications instead of silently swallowed
Fixes#6186
* fix(widget-secrets): add missing drizzle snapshots for merge-url and widget_secret migrations
The merge_custom_widget_url migration (0039/0041/0007) was committed
without a corresponding snapshot, breaking drizzle-kit generate for
all future migrations. This adds:
- 0039/0041/0007 snapshots reflecting the merge-url changes
(base_url+endpoint -> url, add enabled column)
- 0040/0042/0008 snapshots reflecting the widget_secret table
All snapshot chains now match journal entries (41/43/9 entries each).
The refactor from integration-based to option-based config changed the
useQuery enabled condition from 'integrationIds.length > 0' (always true
when an integration was linked) to 'query.length > 0'. The combination
of the debounced search value and the enabled toggle inside the nested
useOptions callback caused the query to never fire, making stations
unsearchable in the edit form.
Removing the enabled guard lets the query always run. The 300ms debounce
already prevents excessive requests.
* docs(helm): Update Helm chart documentation
* feat(widgets): add right-click context menu with auto-generated toggle actions
- Add WidgetContextMenu component using Mantine 9.3 Menu.ContextMenu
- Extend WidgetDefinition with contextActions API for widget-specific actions
- Auto-generate context menu toggle items for switch-type widget options
- Add 'Search' context action to media-requests widgets (list + stats)
- Keep existing dots-button menu for discoverability in edit mode
- Add widgetStateRef to WidgetComponentProps for client-side-only state
- Add skipContextMenu opt-out on switch option definitions
* style: auto-fix lint and format
* fix(widgets): address CodeRabbit review — badge in context menu target, edit-only pending gate
- Move title badge inside WidgetContextMenu wrapper so right-clicking the title overlay also triggers the context menu
- Remove full-menu hide during integration loading; only disable 'Edit' action while isPending
- Add isPending to allActions useMemo dependency array
* style: auto-fix lint and format
* fix(widgets): gate context menu behind authenticated session
- Only show right-click context menu for logged-in users
- Unauthenticated users see plain children without menu wrapper
- Matches existing BoardItemMenu session gating pattern
* style: auto-fix lint and format
* chore: revert unrelated sitemap change from auto-fix
* style: auto-fix lint and format
* fix(context-menu): persist widget option changes to server
Toggle switches and edit modal changes from the context menu now call
saveBoard mutation to persist to DB. Previously only the client cache
was updated, losing changes on reload. Removes unnecessary invalidation
that caused cascading refetch loops.
* style: auto-fix lint and format
* fix: use explicit React type imports instead of global namespace
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
CodeRabbit review addressed:
- Removed the env-var gating (simpler: just remove source maps entirely)
- CI health check now fails fast if container never becomes healthy
Changes vs origin/main:
- next.config.ts: remove productionBrowserSourceMaps and serverSourceMaps
(they default to false, restoring pre-v1.68.0 behavior)
- Dockerfile: remove .map copy step and NODE_OPTIONS=--enable-source-maps
- CI: add memory-regression job (<700MB guardrail)
Closes#6123
Production source maps (--enable-source-maps) caused a ~50% memory increase
(600MB to 1000MB) in v1.68.0. Node.js loads all .map files into memory when
this flag is enabled, and Next.js server chunks generate hundreds of map files.
Changes:
- Guard productionBrowserSourceMaps and serverSourceMaps behind
HOMARR_ENABLE_SOURCE_MAPS env var (default off)
- Guard .map copy step and NODE_OPTIONS in Dockerfile behind
ENABLE_SOURCE_MAPS build arg (default false)
- run.sh checks HOMARR_ENABLE_SOURCE_MAPS before setting --enable-source-maps
- Add Docker-based memory regression CI job (<700MB threshold)
Closes#6123
* Add 'Columns to show' label in Docker stats
* feat: Add column translation keys for Docker widget
* feat: Enhance table with column visibility state
Added column visibility state management for the table.
* Add 'noColumns' error message to en.json
- Paginate GitHub releases via api.paginate to avoid missing tags beyond page 1
- Use tag_name instead of name for GitLab releases (name is a display label)
- Remove ArchiveTeam Warrior from widget-configured note in docs
- Fix test assertion to use objectContaining for error shape with message
- Wrap setSecret delete+insert in a cross-DB transaction
- Add MCP metadata to all widget secret procedures and register in mcp.ts
- Show configured tokens even when no repos use that provider
- Only clear token input after mutation succeeds, with loading state
Every error now includes a diagnostic message:
- noMatchingVersion: shows the regex used, how many tags were
checked, and sample tag names so users can see why matching failed
- noReleasesFound: distinguishes "no releases at all" from "no match"
- invalidIdentifier: shows the problematic identifier string
- unexpected: includes HTTP status codes from provider APIs
getLatestRelease returns { release } | { noMatch: reason } instead
of Release | null, carrying the diagnostic through.
Wrap each repository fetch in try/catch so one failure (Redis,
network, etc.) doesn't crash the entire query batch. Failed repos
now return a proper error response instead of causing all repos
to show "No response from provider".
The releases widget was crashing because throwIfActionForbiddenAsync
or the widgetSecrets query would throw before DB migrations are run.
Wrap in try/catch so the widget works without tokens in that case.
- Fix npm scoped package support (filter time metadata keys, make
description optional, exclude falsy releases from version selection)
- Fix timetable date formatting (use Intl.DateTimeFormat for stable
MM/DD/YYYY output) and named zod import
- Guard migration 0000 against unknown provider keys
- Fix migration 0003 to delete integration links by integrationId
- Reset immich album carousel index when photo list changes
- Add bounded limit to timetable router input
- Remove stale misspelled noProviderSeleceted translation key
- Fix provider name visibility in release widget settings
- Update docs: add ArchiveTeam Warrior note, fix typos, correct
showPhotoInfo default
The widget-only integrations refactor incorrectly removed the
ArchiveTeam Warrior integration entirely. It only needed to be
excluded from auto-seeding (by removing defaultUrl), not deleted.
This restores the integration class, reverts the widget/router/handler
to use integration middleware, and removes archiveTeamWarrior from the
widget-only-integrations migration.
The MantineReactTable defaults table cell backgrounds to
var(--mantine-color-body) which is opaque. This makes the
downloads widget table block appear solid while the footer
and other widgets correctly show the board's background.
Added a styles.css targeting table elements to force
background-color: transparent, following the same pattern
used by the beszel-system-table widget. Also added
bg=transparent to mantinePaperProps for the Paper wrapper.
Adds a Docker button to the user avatar dropdown menu (admin-only)
that opens a 1400px modal containing the same Docker container table
as /manage/tools/docker.
Closes#2165
* fix(manage/about): bypass self-fetch to prevent ECONNREFUSED crash
The /manage/about page fetched its own API routes over HTTP
(http://$HOSTNAME:3000/api/about/...). When the server could not
reach itself (misconfigured HOSTNAME, restricted container networking,
reverse-proxy setups), fetch threw ECONNREFUSED and the whole page
crashed with a full-page Next.js error (issue #6008).
All four routes were force-static serving static JSON / a direct fs
read, so the HTTP round-trip was unnecessary. Replace the fetches
with direct JSON imports and a direct getDependenciesAsync() call,
eliminating the self-network dependency entirely.
Also add a manage-level error.tsx boundary (Next.js 16 convention) so
that any future runtime error in a manage page renders a friendly,
localised fallback inside the existing layout (sidebar stays usable)
instead of the ugly default full-page error.
Exploration draft by an LLM.
Refs #6008
* style: auto-fix lint and format
* refactor: use @static-data path alias, simplify error ternary
- Add @static-data/* tsconfig path alias to replace 7-level relative
imports (../../../../../../../static-data/...) with clean alias
imports across the about page and all 3 API route handlers.
- Simplify error.tsx: replace ternary with && short-circuit.
- Preserve CI lint bot's .sort() -> .toSorted() fix.
Refs #6008
Beszel only stores stats at 1m, 10m, 20m, 120m, 480m intervals.
The 1440m type doesn't exist, so queries for 1w/30d returned empty
results. Use 480m (8h) with appropriate perPage (21 for 1w, 90 for 30d).
* fix(beszel): harden cache layer and prevent blank widgets
- Add debug logging across Beszel integration, request handler, and tRPC
router (auth, fetch, SSE, per-procedure entry/exit) gated behind
LOG_LEVEL=debug
- cached-request-handler: retry with backoff, isValid predicate to skip
caching invalid data, in-flight dedup (single-flight per channel) to
avoid thundering herd after cache invalidation
- redis channel: set-then-publish so storage is always consistent even
when pubsub publish fails
- Beszel handlers: opt into fallbackToStaleOnError and retry so brief
upstream hiccups don't throw away good cached data
- beszel-system-stats: add refetchInterval matching each time period's
cache TTL so historical charts refresh in the background
- tRPC router: switch Promise.all -> Promise.allSettled for getSystems
and getAlerts; getSystemStats returns an error object instead of
throwing. Each result carries an optional error field so one failed
integration no longer nukes the whole widget
- Widgets: add explicit isPending loader, empty state for grid and
stats, and a shared per-integration error indicator (tooltip with
failed integration names). Alerts and table now only show their
empty state after loading completes
- Translation: add empty.noSystems key to beszelSystemGrid and
beszelSystemStats
Addresses #6030
* style: auto-fix lint and format
---------
Co-authored-by: ajnart <ajnart@users.noreply.github.com>
* feat(analytics): migrate from Umami to PostHog server-side analytics
Replace @umami/node with posthog-node for all server-side telemetry.
Add singleton PostHog client, feature tracking POC, DAU heartbeat,
and tRPC analytics router. Cron job is now locked (no manual trigger,
no custom interval). Update consent copy to explain what we collect
and why.
* style: auto-fix lint and format
* refactor(analytics): simplify + add client-side posthog-js init
- Inline constants.ts and track.ts into client.ts (single source)
- Remove dead exports (shutdownPostHogAsync, getPostHogClient)
- Flatten all DB queries into one Promise.all (no waterfall)
- Derive integration/widget totals from group-by (drop redundant counts)
- Add posthog-js with dynamic import, gated on enableGeneral
- Move Analytics component to body, pass consent from server settings
* style: auto-fix lint and format
* feat(cron): add hidden flag to cron jobs, hide analytics from task panel
The analytics job runs silently in the background — users don't need
to see or interact with it. Hidden jobs are filtered out of getAllAsync
so they never appear in the management UI.
* refactor(analytics): simplify to single toggle, always collect all data
Remove granular enableWidgetData/enableIntegrationData/enableUserData
toggles. Analytics is now a single on/off switch. When enabled, all
anonymous usage data is collected in one flat Promise.all call.
Side effect: existing users with enableGeneral=true who had granular
flags off will now send all data. This is by design — the granular
controls added complexity without meaningful privacy benefit since
the data is already anonymous counts.
* style: auto-fix lint and format
* Calculate and display total CPU and memory usage
Added total CPU and memory usage calculations for containers.
* Refactor Docker component layout
Refactor component layout and remove redundant Group elements.
* Create test.yaml
* Add files via upload
* Delete packages/icons/icons directory
Accidental
* Add total CPU and memory strings to English translations
* Add English translations and size refactor
* Fix typo in search label in en.json
* chore(docs): update OpenAPI spec
* Remove color styling from CPU and memory usage text
* Enhance Docker component with footer and toolbar updates
Updated createColumns function to include totals and relativeTime parameters for footer display. Adjusted table configuration to enable bottom toolbar based on isTiny flag.
* Fix formatting and add missing newline at end of file
* Enable sticky header and footer in component
Enable sticky header and add footer row properties for sticky positioning.
* Refactor table component properties
* Refactor table columns and fix footers
* Revert "chore(docs): update OpenAPI spec"
This reverts commit 26f3c809d9.
---------
Co-authored-by: Bergasha <Bergasha@users.noreply.github.com>
Co-authored-by: Manuel <30572287+manuel-rw@users.noreply.github.com>
Stalled torrents have time=0 (infinite/never), which raw numeric sort
places between negative (completed) and positive (active) values.
Add a custom sortingFn on the time column that maps 0 to Infinity so
stalled entries always sort last in ascending order and first in
descending order.
Closes#5416
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
- Remove ternary operators from new code (widget-integration-select,
TabsContainer, mock data, members page)
- Remove stale /docs/management/tools/backup from sitemap type
- Add value prop to error-state Select in beszel-system-stats
- Add generation guard to prevent late SSE events from writing stale
data after systemId/integration changes
- Sort integrationIds for stable key (avoids unnecessary resubscribes)
- Include integrationKey and systemId in flush interval deps
- Preserve current form value in WidgetIntegrationSelectInput when it's
missing from loaded options (matches WidgetDynamicSelectInput pattern)
react-jsx-parser auto-generates random keys on every render, causing
full remounts that reset uncontrolled Tabs to defaultValue. Fix by:
- Adding disableKeyGeneration to preserve React reconciliation
- Memoizing bindings to avoid unnecessary re-parses
- Converting TabsContainer to controlled Tabs with useState
- Validating defaultTab against actual panel values
- Clearing stale parseErrors when template/data changes
The Gotify integration (merged from dev) declares a documentationUrl but
had no docs page, failing the docs-coverage validation plugin once gotify
was added to the integration doc-slug map.
Merging dev introduced the Gotify integration and ArchiveTeam Warrior
widget/integration, which require entries in the doc-slug maps and the
generated documentation sitemap. Also remove a misplaced non-null
assertion in the group members page flagged by oxlint.
- Throw TRPCError instead of returning null when no integration
- Sort beszelRouter import alphabetically in mcp.ts
- Replace ternary with lookup table for background color
- Mention integration_all in MCP server instructions
- Add .default("1h") to timePeriod so MCP clients can omit it
- Include "1m" in the description's valid timePeriod values
- Clarify that getSystemStats uses only the first integrationId
- Use MantineSize type instead of plain string for itemRadius
Register beszelRouter in the MCP server with .meta({ mcp: enabled })
on getSystems, getAlerts, and getSystemStats procedures. AI assistants
can now query Beszel-monitored systems, view alerts, and fetch
historical metrics through the MCP protocol.
Ref #5968
The inner Card components used Mantine's default opaque background and
hard-coded radius, clashing with board styling. Now uses board.opacity
and board.itemRadius with the same pattern as system-resources widgets.
Closes#5968
* feat(integrations): add Gotify notifications integration
Closes#4186
> "It would be really useful if Homarr could integrate with Gotify —
> view notifications from Gotify right inside the Homarr dashboard,
> maybe as a widget or a small panel. A widget that displays the latest
> messages from Gotify (with title, message, time)."
> — #4186: feat: Gotify Integration - display notifications in Homarr
- `gotify/gotify-schema.ts` — Zod schema validating the `GET /message`
API response from the Gotify server, including id, date, title,
message body, and optional priority fields.
- `gotify/gotify-integration.ts` — `GotifyIntegration` class extending
the base `Integration` and implementing `INotificationsIntegration`.
Authenticates via HTTP Basic Auth (username + password) and fetches
the latest 100 messages from the `/message` endpoint. Connection
testing hits `/health`.
- `index.ts` — exports `GotifyIntegration` from the package public API.
- `base/creator.ts` — registers `GotifyIntegration` in the integration
creator map so it can be instantiated at runtime.
- `integration.ts` — registers the `gotify` integration definition with:
- `secretKinds: [["username", "password"]]` (Basic Auth credentials)
- `category: ["notifications"]` (picked up automatically by the
Notifications widget via `getIntegrationKindsByCategory`)
- Icon sourced from `homarr-labs/dashboard-icons`
- Documentation URL pointing to `/docs/integrations/gotify`
- `docs/homarr-docs-sitemap.ts` — adds `/docs/integrations/gotify` to
the `HomarrDocumentationPath` union type so the documentation link
passes the TypeScript path validation check.
Because `gotify` is registered under the `notifications` category, the
existing Notifications widget automatically supports it with no widget
code changes required. Users add the integration via Settings →
Integrations, supply their Gotify server URL, username, and password,
then attach it to a Notifications widget on any board.
* refactor(integrations): inline json parse and add error handling in Gotify integration
Address PR review feedback: remove intermediate `as unknown` cast by inlining
response.json() directly, and use safeParseAsync to handle deserialization
failures gracefully instead of throwing unhandled.
On macOS (and some Glances configurations) the /api/4/all endpoint
omits the quicklook key, causing a Zod invalid_type error and crashing
the system health widget. Mark quicklook as .optional() in allSchema
and add a null-safe fallback (`?? "Unknown"`) for cpuModelName.
Fixes#5728
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Consolidate 5 stat mapper useCallbacks and 6 series useMemos into
single mappers/series objects. Extract statsWhenShown helper for
repeated visibility guard pattern. Deduplicate SSE record envelope
construction. Simplify prepareRecords with if/else. Remove unused
gridColumns ternary with Number() indexing.
Replace all ternary operators with if/else, &&/||, or early returns.
Remove unused noAnimation constant that caused lint error.
Fix variable shadowing in tooltip deduplication loop.
- Decouple 1Hz SSE ingestion from render rate: accumulate data in a
ref and flush to React state every 3s instead of every tick
- First data point flushes immediately to avoid blank screen
- Reduce MAX_BUFFER from 60 to 30 (halves SVG reconciliation work)
- Gate chart data hooks behind visibility flags so disabled charts
skip all array processing
Keep showing existing charts during brief WebSocket reconnections
instead of flashing the error UI. The error state only shows when
no data was ever received (genuine initial connection failure).
- Set withDots={false} to eliminate thousands of hidden SVG Dot nodes
- Disable Recharts animation (isAnimationActive: false) during data updates
- Merge ChartPanel + BeszelAreaChart into BeszelChartPanel to fix
React.memo being defeated by children prop creating new refs
- Memoize yAxisProps inside BeszelAreaChart instead of rebuilding inline
- Hoist chartStyle, panelStyle, CPU_Y_AXIS_DOMAIN to module constants
- Memoize subscription input/enabled in useLiveStats
- Replace O(n) find() with Map lookup in useDockerChartData hot path
- Stabilize useContainerNames with ref comparison so it returns the
same array reference when container names haven't changed
- Memoize containerSeries (was computed inline every render)
- Hoist all system chart series arrays into useMemo
- Hoist activeDotProps/dotProps to module-level constants
- Wrap BeszelAreaChart and ChartPanel in React.memo
When SSE connection fails, the error UI now shows a specific message
and a button to switch back to historical view, so users don't need
to reload the page to recover.
- Restore timePeriod union cast to fix typecheck (TS2769)
- Clear live data on subscription error so error UI always shows
- Emit error when no integration is configured for live subscription
- Add error state to system select input when query fails
- Surface subscription errors in useLiveStats via onError + error state
- Show error UI in live mode when SSE connection fails
- Abort SSE reader on subscribeTopic failure instead of silent log
- Guard mock realtime against pre-aborted signal to prevent interval leak
- Remove dead includeDocker from subscription input schema
- Gate stats/live queries on systems loading state to prevent stale ID
requests during initial load
- Simplify useLiveStats: functional setState, stable integrationKey dep,
remove unused includeDocker param
- Extract prepareRecords helper to deduplicate chart ordering logic
- Remove stale timePeriod type cast
Replace plain text systemId option with a dedicated beszelSystem factory
type backed by a searchable Select that fetches available systems from
the integration. Shows proper states for no integration, loading, and
no systems found.
- Skip toReversed() for live data (already chronological from buffer)
- Reset live buffer on systemId/integration change to avoid stale data
- Throw on unexpected SSE stream end so subscription emits error
- Delete unused time-period.ts config (duplicate of server-side config)
Add "Live" option to time period selector that uses the existing
subscribeSystemStats tRPC subscription (PocketBase SSE) instead of
REST polling. A useLiveStats hook accumulates incoming data points
in a 60-entry rolling buffer. Charts use HH:mm:ss timestamps in
live mode.
- Add subscribeSystemStats tRPC subscription bridging PocketBase SSE
- Add subscribeRealtimeMetrics to BeszelIntegration with SSE parsing
- Add error handling: try/catch in subscription, response check in topic POST
- Add cacheDurationForInput to request handler for per-period cache TTL
- Add mock subscribeRealtimeMetrics with 1s interval for dev testing
Non-stacked charts produce duplicate Recharts payload entries (stroke +
fill). Deduplicate by dataKey keeping the first entry (correct color).
Also increase indicator borderRadius for a fully rounded pill.
BeszelAreaChart hardcoded type="stacked" for all charts, causing Docker
Network I/O to render blended colors that didn't match the tooltip.
- Make `type` prop configurable (default "default")
- Only stack Memory, Docker CPU, and Docker Memory charts
- Add systemId as a persisted widget option with Menu-based selector
- Replace Popover with Menu trigger="click-hover" for proper hover UX
- Show error state with Select when selected system is not found
- Remove 1-minute time period (live SSE not wired yet)
The SWC HTML minifier rejects slightly invalid HTML that browsers
tolerate, breaking the build on every page. Disable it while keeping
all other Docusaurus Faster optimizations.
Also fix broken relative links in after-the-installation.mdx and
donate.mdx by using absolute paths.
GitHub Pages redirects non-trailing-slash URLs with 301, causing
Algolia crawler to ignore all sitemap entries. Setting trailingSlash
to true aligns generated URLs with what GitHub Pages actually serves.
The postinstall codegen script crashes CI when homarr.dev/sitemap.xml
is unreachable. Now catches fetch errors and keeps the existing
committed homarr-docs-sitemap.ts instead of crashing.
* chore(deps): update dependency proxmox-api>undici to v7.27.1
* style: auto-fix lint and format
* chore: also update undici
---------
Co-authored-by: homarr-renovate[bot] <158783068+homarr-renovate[bot]@users.noreply.github.com>
Co-authored-by: Manuel <30572287+manuel-rw@users.noreply.github.com>
* fix(boards): add permissions to board summary schema to fix manage page crash
The getAllBoards query fetches userPermissions and groupPermissions
from the DB, but boardSummarySchema (the output validator) did not
include them. tRPC stripped the fields, causing a TypeError when
getBoardPermissionsAsync called .some() on undefined.
* fix(backup): cast Buffer to Uint8Array for NextResponse body
Node 24 narrowed the BodyInit type, making Buffer incompatible.
Wrapping in Uint8Array satisfies the type constraint.
* fix(lint): remove forbidden non-null assertions across backup and board code
Replace non-null assertions with safe guards, optional chaining, and
for-of iteration so oxlint passes in CI. Also remove unused Collapse import.
`import.meta.url` is undefined when esbuild bundles migrate.ts into
migrate.cjs, causing a TypeError at startup. Inline the seed JSON
data directly instead of reading from disk, which also fixes the
Docker image missing the seed files (not copied by Dockerfile).
* chore(integrations): add ws dependency for the truenas client
* feat(definitions): allow api key authentication for truenas
* fix(integrations): connect to truenas over trusted-certificate websocket with protocol autodetection
* test(integrations): cover truenas protocol detection and authentication
* refactor(integrations): avoid shorthand index variable in truenas mapping
* feat(integrations): add a websocket url helper to the integration base
* refactor(integrations): extract truenas connection handling into a client
createTheme was called without being imported, causing a runtime
ReferenceError on every page load. Use the already-imported theme
object from @homarr/ui which contains the full Mantine configuration.
* feat(board): persist category collapse state in localStorage for guests
Authenticated users already had their collapse state saved server-side.
This extends the same behaviour to unauthenticated (guest) users by
reading/writing localStorage keyed by section ID.
On mount, if the user is a guest, the stored value is applied to the
disclosure. On toggle, guests write to localStorage instead of calling
the protected API mutation.
Closes#3355
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* refactor: set localStorage value as useDisclosure default instead of useEffect
Replaces the useEffect that read localStorage after mount with a direct
initial value computation passed to useDisclosure. Removes the effect,
the open/close destructuring, and the useEffect import.
Note: initialCollapsed is computed during render rather than post-mount,
which may cause a hydration mismatch on SSR if localStorage differs from
the server-provided section.collapsed. The typeof window guard handles
the SSR case.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
When adding a widget from the "Choose item to add" modal, the widget
edit modal now opens immediately after placement, allowing the user to
configure options and integrations before the widget renders.
All available integrations for the widget are selected by default.
* ci: add --affected filtering and fix Turbo cache invalidation
- Remove globalDependencies **/.env which invalidated all caches on any .env change
- Move 40+ env vars from globalEnv to build-specific env (lint/format/typecheck don't need DB_URL etc.)
- Add per-task inputs with test file exclusions so test changes don't trigger build/typecheck
- Enable affectedUsingTaskInputs for task-level --affected detection
- Add --affected to all turbo commands in CI (lint, format, typecheck, build)
- Add fetch-depth: 0 for proper git history comparison
- Add Docker Buildx + GHA layer caching for E2E builds
* ci: add Cloudflare remote cache for Turborepo
Configure turborepo-remote-cache-cloudflare as the remote cache backend.
This commit can be reverted independently to disable remote caching.
* fix(ci): set TURBO_SCM_BASE for --affected to work in PRs
Turbo defaults to comparing against `main`, but PRs target `dev`.
Without explicit base/head refs, --affected falls back to running
all tasks. Use the PR base SHA for pull_requests, the before SHA
for pushes, and HEAD~1 as a last resort.
* perf(typecheck): exclude bundled .cjs/.css from tsc
websocket (835K lines) and tasks (755K lines) had bundled output
files being typechecked due to `include: ["."]` in their tsconfig.
Full typecheck drops from ~70s to ~26s.
* perf(turbo): scope topo task inputs to package.json only
The topo virtual task had no explicit inputs, defaulting to all files.
This caused any source change to cascade cache invalidation through
all dependent packages via ^topo. Since topo is only for dependency
ordering, its hash should only change when dependencies change.
* fix(turbo): disable remote cache signature verification
The cache lives on a private CF worker, signature verification
just adds noise without security benefit.
* fix(turbo): remove runtime env vars from build cache key
These env vars (DB_*, AUTH_*, DOCKER_*, etc.) are all runtime
configuration provided by users when running the Docker container.
They don't influence the build output since Next.js standalone mode
reads process.env at runtime and SKIP_ENV_VALIDATION=true skips
validation during build. Listing them in build.env caused unnecessary
cache busting in local dev.
Also restores coverage report comments per review feedback.
* fix(turbo): include test files in typecheck cache inputs
Test files must be typechecked — excluding them from typecheck inputs
caused changed tests to skip type validation.
* fix(deps): update dependency zod to ^4.3.5
* fix: omit can not be used with refine
* fix: convert intersection to object due to trpc-openapi issue
* fix: remove omit after refine
---------
Co-authored-by: homarr-renovate[bot] <158783068+homarr-renovate[bot]@users.noreply.github.com>
Co-authored-by: Meier Lukas <meierschlumpf@gmail.com>
* fix(deps): update dependency node-ical to ^0.23.1
* fix: date from node-ical represents utc value
---------
Co-authored-by: homarr-renovate[bot] <158783068+homarr-renovate[bot]@users.noreply.github.com>
Co-authored-by: Meier Lukas <meierschlumpf@gmail.com>
* feat: add GPU monitoring support via Glances integration
- Add GPU fields to SystemHealthMonitoring interface (gpuId, name,
memoryUtilization, processorUtilization, temperature, fanSpeed)
- Extend Glances integration to parse GPU data from /api/4/all endpoint
(supports NVIDIA, AMD, and Intel GPUs via Glances)
- Add GPU utilization ring to Health Monitoring widget with tooltip
showing VRAM usage, temperature, and fan speed
- Add GPU time-series chart to System Resources widget (grape color)
- Add GPU toggle option to both widgets
- Add English translations for GPU labels
- Add gpu: [] to all other integrations (dashdot, truenas, unraid,
openmediavault) for backward compatibility
- Add GPU mock data for development/testing
Closes community request from PR #5102 for GPU monitoring support.
* fix: implement GPU monitoring for dashdot integration and clarify OMV comment
- Add GPU load fetching via dashdot /load/gpu endpoint
- Add GPU info parsing from dashdot /info endpoint
- Add gpuLoadApi and gpu field to internalServerInfoApi schemas
- Add getCurrentGpuLoadAsync method to DashDotIntegration
- Map GPU load data to SystemHealthMonitoring gpu array
- Add explanatory comment for OMV's empty gpu array
Addresses review comments from manuel-rw on PR #5255
* Delete manifest.json
When Docker API returns `null` for container ports (instead of `[]`),
the Docker management page crashes. Added optional chaining on
`ports?.length` to handle this case.
Fixes#5311
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
* feat(analytics): consolidate server analytics into single Umami event
- Replace multiple separate events with single 'server-analytics' event
- Add comprehensive metrics: Homarr version, database type, Docker/K8s status
- Add auth providers status (credentials, OIDC, LDAP)
- Add counts for users, boards, groups, apps, widgets, sections, integrations
- Add counts for search engines, icon repos, icons, API keys, invites, media
- Add counts for layouts, item layouts, section layouts, cron job configs
- Include per-integration-kind breakdown (integration_sonarr, etc.)
- Add required dependencies: @homarr/auth, @homarr/common, @homarr/docker
* refactor(ui): simplify analytics settings to single toggle
- Remove individual toggles for widget/integration/user data
- Keep only the main 'Enable analytics' toggle
- All data is now collected when analytics are enabled
- Update both init settings and manage settings pages
* docs(i18n): update analytics description with collected data details
- Explain all data points collected when analytics are enabled
- List: version, DB type, Docker/K8s, auth methods, counts
- Clarify that no personal info, URLs, or names are collected
* chore(deps): update pnpm-lock.yaml for analytics package dependencies
- Add 'orphaned' container state to handle containers with null/undefined cpu_stats.online_cpus
- Display orphaned containers with gray badge in UI
- Prevent 'Cannot read properties of undefined (reading online_cpus)' error
- Add translations for 'orphaned' status in all supported languages
- Containers with invalid stats are now marked as orphaned instead of causing errors
description:Navigate Homarr's monorepo architecture and reuse shared packages. Use when locating code, choosing a package boundary, understanding runtime services, tRPC, databases, widgets, integrations, cron jobs, Custom Widgets, Workshop, or checking @homarr/common before creating a utility.
---
# Homarr Codebase Context
Orient from the current checkout before editing. Treat package manifests, exports, and source as authoritative when they differ from this snapshot.
## Workflow
1. Confirm the branch and inspect the nearest `package.json`, existing sibling modules, and package exports.
2. Search `@homarr/common` before creating a general-purpose helper. Read [references/common.md](references/common.md) for every public entrypoint and export.
3. Read [references/architecture.md](references/architecture.md) when the task crosses packages, runtimes, routing, data, widgets, integrations, cron jobs, Custom Widgets, or Workshop.
4. Keep the change in the deepest package that owns the behavior. Reuse public entrypoints rather than deep-importing implementation files.
5. Re-run targeted searches before documenting counts, consumers, or dependency edges; these change faster than the architectural seams.
## Current toolchain
- Use Node `24.18.0` for application runtimes and Bun `1.4.2` for package management, pinned in `mise.toml` and `package.json`.
- Use Bun workspaces with Turborepo and a shared hoisted install. Workspace packages use the `@homarr/` scope and catalog-managed dependencies.
- Use Next.js App Router, TypeScript, tRPC, Drizzle, Mantine v9, Tabler icons, Jotai, TanStack Query, next-intl, Vitest, and Playwright.
- Use oxlint and oxfmt. Mantine is the application UI system; Tailwind is limited to the docs app.
-`bun run dev` starts only `@homarr/nextjs`. Start a standalone runtime explicitly when the task needs it.
Invoke `documentation-sync` only when users need new or corrected information to set up, use, or troubleshoot Homarr; follow the Documentation Sync criteria in `AGENTS.md`. When exposing tRPC through MCP, invoke `mcp-integration`.
| `apps/nextjs` | Main Next.js application and HTTP APIs on port 3000 |
| `apps/docs` | Next.js and Fumadocs documentation site on port 3003 |
| `apps/tasks` | Cron-job initialization and scheduling runtime; it does not expose the old port-3002 API |
| `apps/websocket` | tRPC WebSocket subscriptions on port 3001 |
| `apps/workshop` | Go/PocketBase Workshop service and provider integration |
`bun run dev` runs only `@homarr/nextjs`. Use `bun run --filter @homarr/tasks dev`, `bun run --filter @homarr/websocket dev`, or `bun run dev:docs` when a task needs those runtimes.
In production, `apps/nextjs/src/instrumentation-node.ts` starts the tasks and WebSocket runtimes inside the Bun process. Nginx listens on port 7575, proxies `/websockets` to port 3001, and sends other traffic to port 3000. Redis can run internally or externally.
## Package map
### Foundation and domain
-`core`: environment validation, database and Redis factories, logging, infrastructure.
-`widgets`: widget definitions, dynamic component loaders, and shared widget UI.
-`translation`: next-intl configuration and locale catalogs.
-`cli`: Bun operations CLI. `tools/homarr-dev` is the separate Go developer CLI.
## Dependency direction
Prefer these directions; verify the exact dependency in package manifests before changing a boundary.
1. Put infrastructure in `core` and stable domain names in `definitions`.
2. Put reusable, dependency-light helpers in `common`; use its public environment-specific entrypoints.
3. Build persistence in `db`, identity in `auth`, transport in `redis`, and external clients in `integrations`.
4. Compose server behavior in `api`, `cron-jobs`, and request-handler packages.
5. Build primitives in `ui`, `form`, and `modals`; compose product surfaces in widgets and collection packages.
6. Keep app entrypoints in `apps/*` thin. Avoid moving reusable behavior into `apps/nextjs` when a package already owns it.
## Core patterns
### tRPC and MCP
`packages/api/src/root.ts` lazily composes the application router. Server components use `@homarr/api/server`; client components use `@homarr/api/client`. MCP extraction requires the eager router in `packages/api/src/mcp.ts`, so MCP-enabled subrouters must be imported and registered there directly. Use the `mcp-integration` skill for exposure and security rules.
### Databases
Drizzle schemas live in `packages/db/schema/`. `DB_DRIVER` selects `better-sqlite3` or `node-postgres`; MySQL and MariaDB are no longer runtime drivers in v2. Existing MySQL installations must use the documented converter before upgrading. Migrations live under `packages/db/migrations/sqlite/` and `packages/db/migrations/postgresql/`.
### Widgets
`packages/widgets/src/registry.ts` owns the explicit widget-module and component-loader registries. `packages/widgets/src/manifest.ts` caches those loaders and exposes definitions, components, and combined resources. Widget modules use `createWidgetDefinition` and typed option builders. Data commonly comes from tRPC queries or Redis-backed cron results. Use existing shared widget components before creating local variants.
### Integrations
`packages/definitions/src/integration.ts` owns `integrationDefs` and `IntegrationKind`. `packages/integrations/src/base/creator.ts` owns `createIntegrationAsync`; concrete integrations implement the shared base contract. Keep credentials and permission checks on server-owned paths.
### Cron jobs
Jobs live in `packages/cron-jobs/src/jobs/` and use `createCronJob`. `apps/tasks` initializes and starts the job group. Job results and status can flow through Redis and tRPC; the retired port-3002 management-API model is no longer part of the architecture.
### Routing and localization
User routes live below `apps/nextjs/src/app/[locale]/`; the locale segment does not add a URL prefix. Route groups organize home and board surfaces. Administrative routes live under `manage`, Custom Widget configuration has a token route, and Workshop has a management surface.
### Custom Widgets and Workshop
`packages/custom-widgets` owns the Custom JSX v2 contract. `packages/workshop` is the Homarr-side client/contract package; `apps/workshop` is the separate Go/PocketBase service. Use the `homarr-custom-widget` skill for authoring and security constraints.
## Commands and conventions
- Read current scripts from the root or package `package.json`; avoid caching commands that do not exist.
- Use `bun run dev` for Next.js, `bun run dev:docs` for docs, and `bun run dev:cli -- <args>` for the Go developer CLI.
- Use `bun run docker:dev:up` for the Redis development dependency.
- Follow the Testing section in `AGENTS.md` for test creation and validation scope.
- Use `~/*` only inside the Next.js app. Use public `@homarr/*` entrypoints across package boundaries.
- Keep Mantine application styles separate from the Tailwind-based docs app.
Import only from an exported package entrypoint. Do not deep-import `@homarr/common/src/*`. Check `packages/common/package.json` and the source again when adding or renaming an export; this inventory describes the current repository tree.
`NodeFetchHttpErrorHandler` exists as an implementation file but is not exported by the public server entrypoint. Add an intentional public export before using it across packages.
## Client and environment exports
-`@homarr/common/client` exports `revalidatePathActionAsync`, a Next server action for page revalidation.
-`@homarr/common/env` exports `env` with validated `NODE_ENV`, `SECRET_ENCRYPTION_KEY`, and `NO_EXTERNAL_CONNECTION` values.
description:Write Homarr docs only when users need information to set up, use, or troubleshoot the app. Skip routine, obvious, or self-explanatory changes.
---
# Documentation Sync
Apply the audience and scope rules in `AGENTS.md` first. Documentation exists only for information users need to set up, use, or troubleshoot Homarr. A code, UI, or API change alone does not warrant documentation; if a user does not need to know it, do not add it.
For hidden capabilities such as the advanced widget feature, explain discovery, activation, and non-obvious behavior. For visible standard controls such as Delete, omit narration of what the label already conveys. Use the shortest explanation that preserves necessary technical detail; remove redundant prose in the passage being edited.
When an update is warranted, inspect adjacent pages for structure and types, not as a verbosity target. The mappings below locate needed information; they do not require a page for every feature or code change.
## Locate a needed update
These are destinations for warranted updates, not triggers to add documentation:
| Permission behavior | `apps/docs/docs/management/users.mdx` and the affected feature page |
| Custom Widget or Workshop behavior | Relevant Custom Widget, Assistant, or Workshop documentation |
If no existing row fits, search by the user-visible term and update the page where a user would look for the behavior:
```bash
rg -n "<term>" apps/docs/docs apps/docs/src
```
## Follow existing typed patterns
For an integration:
- Add or update `apps/docs/docs/integrations/<slug>/index.ts` with the local `IntegrationDefinition` pattern.
- Add or update `index.mdx` with the established `IntegrationHeader`, `IntegrationCapabilites`, and `IntegrationSecrets` components as applicable.
- Reuse metadata from `@homarr/definitions` when the docs app already exposes it.
For a widget:
- Add or update `apps/docs/docs/widgets/<slug>/index.ts` with the local `WidgetDefinition` pattern.
- Add or update `index.mdx` with the established `WidgetHeader`, `WidgetConfig`, and `AddingWidget` components as applicable.
- Explain only non-obvious option semantics, integration prerequisites, permission constraints, or behavior that the interface does not make clear.
## Completion criteria
1. Every passage adds information a user needs and cannot reasonably infer from the interface. Remove obvious, repetitive, or unnecessary prose.
2. Keep names, defaults, paths, screenshots, links, and prerequisites consistent with code.
3. Update every affected page and remove superseded guidance.
4. Run the narrowest useful docs validation. Use `bun run turbo build --filter=@homarr/docs` when links, MDX, generated definitions, or navigation can fail; otherwise run the docs package formatter on touched files.
5. Treat broken links and anchors as failures; run `bun run --filter @homarr/docs validate:links` after the static build.
Author requested widgets with release context; validate, test, persist, and return artifacts.
- Read primary API docs when missing/changed. Samples and successful previews are binding; load only needed schema, runtime,
security, or component context.
- Batch unknown component searches/details. `contextAlreadyLoaded` reuses context; `phaseComplete` advances. Stop only for
genuine provider/model, lifecycle-service, or workbench-closure failure.
- Community widgets use `customWidget_workshopSearch`, `customWidget_workshopGet`, and
`customWidget_workshopInstall`; configure and persist.
Return one fenced `json` block with the complete definition; keep evidence prose outside it. The definition has keyed
`sources`, `requests`, `template`, and optional `options`; actions are requests with `kind: "action"`.
-`sources.default` is required. HTTP has `baseUrl`, `networkScope`, and credential-free `auth`; localhost/loopback requires
`networkScope: "loopback"`; never widen an explicit scope. Saved sources use `type: "integration"`/`integrationKind`;
Homarr holds credentials.
- Saved integrations: discover kinds/full-access entries with `integration_getKinds`/`integration_all`, bind `integrationId`
before preview, omit URL/auth, and keep non-GET requests as actions.
- Paths are slash-prefixed: strings use `{option:name}`/`{param:name}`; query/body uses `{"$option":"name"}`/`{"$param":"name"}`.
Loads use `trigger: "load"`; manual helpers use `trigger: "manual"`.
- Actions stay manual; preserve `confirmation`, `permission`, and `invalidates`; DELETE requires full permission/confirmation.
`$param` is manual-only; `$option` may drive loads.
- Load data.requestId/status.requestId with RefreshButton; status.requestId?.ok === false is error. Manual SubFetch never
populates data/status; its child receives (result, metadata) and renders its fields.
- Options have `label`, `control`, `default`; installation config is `options.name`, never `inputs`. Request-bound TextInput,
Select, NumberInput, Pagination use literal `bind` + default and manual `SubFetch params` map `inputs.<name>` to `$param`.
Dependent pagination uses `defaultValue={1}`/`resetKey={inputs.query}`. Remove controls without an option/request/helper; guard arrays/nested with `??`;
preserve documented timezone values; use UTC only when the contract says UTC.
- Templates are one expression: no imports, hooks, refs, raw HTML/events, browser requests, eval, recursion, IIFEs,
statement blocks, or arbitrary functions. Use registered component names; `Icon` may alias `TablerIcon`. Keep hierarchy,
theme tokens, useful states, and narrow/wide layouts purposeful.
## Bounded lifecycle
1. Build a credential-free definition from request, verified context, and sample. Preserve a migration's API path, method,
body, options, and behavior; omit unknown requests rather than guessing.
2. Send the coherent complete definition directly to `customWidget_previewCreate`; it validates both manifest and JSX.
Use `customWidget_validateTemplate` only for isolated JSX diagnostics, never as a preview prerequisite. Use
`customWidget_previewReviseTemplate` for JSX-only corrections after a preview exists. In the Assistant wrapper, multiline
JSX uses `templateLines` and preview creation receives the complete definition.
3. Test every returned query/simulated action once, batching independent queries. On a concrete schema/preview error, fix only
that field and retry `customWidget_previewCreate` with the corrected definition; use `customWidget_previewReviseTemplate`
only for JSX errors after a preview exists. Stop only for genuine provider/model, lifecycle-service, or
workbench-closure failure.
4. If `previewCreate` used `definitionId`, persist with `customWidget_updateFromPreview`; otherwise use
`customWidget_createFromPreview`. Follow create `nextAction` once. Configure credentials in Homarr; never repeat plaintext
secrets.
## Delivery
Report actual lifecycle results. If unavailable, add one post-artifact `Unverified:` line naming missing validation, preview,
renderer, or persistence. Never claim rendering/persistence from schema checks.
Load-query templates read `data.requestId` and `status.requestId`, plus `options.name` and temporary `inputs.name`. Manual `SubFetch` results stay local to that instance and never populate `data`/`status`; render them in its child callback. Status is `{ loading, ok, status, statusText, error }`. Render load queries directly from `data` and `status` with `RefreshButton`; never wrap them in `SubFetch`.
`bind` is temporary; manual request values go in `params` and map to `$param`:
Manual queries require `trigger: "manual"` on request and `SubFetch`; otherwise they run automatically. `triggerContent` with `triggerAriaLabel` makes custom content the launcher. `SubFetch` owns loading/error/retry; its child receives success plus `{ ok, status, statusText, loading: false }`. Never author `onClick` or fetch callbacks.
`SubFetch`, `ActionButton`, and `ToggleSwitch` need literal `requestId`; validation rejects missing/computed IDs.
Inside a successful manual result, `<RefreshButton requestId="search" label="Run again" />` reruns the same parameters.
When a manual SubFetch ID, params, or definition changes, Homarr hides the prior result and returns to the trigger; new params fetch only after triggering again.
Exact paths: load request ID `q` with raw preview body `B` -> `data.q === B`. `q=events` body`{"events":[...]}` -> `data.events.events`; do not flatten repeated keys. Manual `SubFetch` receives `B` as result -> `result.events`. Before persistence inspect core paths against preview; revise JSX and retest if mismatched
Format timestamps with safe static helpers; never use `new Date`. Never invent a formatter component. Use `Date.toLocaleString(value, "en-US", documentedTimezone)` and label the documented timezone; if no timezone is documented, preserve the source value or omit any timezone label; use UTC only when the response contract says UTC. Also available: `Date.toISOString`, `Date.toLocaleDateString`, and `Date.toLocaleTimeString`.
For compact numeric enums, index a literal label array with a fallback:
Request-bound controls use literal `bind` plus a default (`defaultChecked` for Switch/Checkbox); pass `inputs.<name>` through manual `SubFetch params` to matching `$param`. Options are installation config via `options.name`, never `inputs`; dependent pagination uses `defaultValue={1}`/`resetKey={inputs.query}`. Remove dead controls.
Callback parameters must not shadow the reserved roots `data`, `status`, `options`, or `inputs`. Use registered component names returned by discovery; `Icon` is an accepted alias for canonical `TablerIcon`. Never invent components such as `<IconFoo />`.
Use expression callbacks for supported collections and trusted slots. No callback blocks, IIFEs, authored recursion, or raw events. Regex is limited to safe string operations.
Key `default` is the required source ID, not a source property. Fields: `name?`, `baseUrl`, `networkScope`, `auth?`; localhost/loopback URLs require `networkScope: "loopback"`; never widen explicit scope:
Saved sources use `{"type":"integration","integrationKind":"sonarr","integrationId":"saved-id"}`. Discover and bind a full-access entry before preview. Omit URL/auth; paths append to its URL, non-GET requests are actions, and exports omit `integrationId`.
Auth is `none`, `bearer`, `basic`, `apiKeyHeader`, or `apiKeyQuery`. Requests default to source `default`, query/GET/load, inherited auth, and view permission. Use `load` for initial/current display and `manual` only for explicit interactions or invocation params. Actions are manual/modify; DELETE requires full permission and confirmation.
JSON responses become their decoded value. Responses with `application/x-ndjson` become an array with one decoded object per non-empty line.
Use real public API URLs and clear self-hosted placeholders. Homarr collects URL, scope, and credentials outside the manifest.
Paths use `{option:name}`/`{param:name}`; query/body objects use `{"$option":"name"}`/`{"$param":"name"}`. `$param` is manual-only; `$option` may drive loads. Constants stay primitive.
Every option has `label`, `control`, and `default`. Optional fields are `description`, `choices`, `choicesFrom`, `min`, `max`, `step`, `advanced`, and `group`.
All requests use Homarr's protected server executor. Source origin, network scope, DNS, redirects, SSRF, rate limits, permissions, size limits, timeouts, and encrypted credential injection remain enforced.
The JSX interpreter blocks imports, hooks, refs, raw event callbacks, browser requests, eval, arbitrary functions, prototype access, unsafe URLs, global CSS escape, arbitrary portals, bigint, statement blocks, IIFEs, and recursion. Regex literals must be bounded and reject backreferences, lookbehind, nested quantifiers, excessive length, and unsupported flags.
Credentials are stored separately and never exported or returned to an agent. A published self-hosted source URL is only a suggestion: installers must confirm or replace private and loopback URLs for their own Homarr deployment. Source origins cannot be controlled through widget options.
description:Expose Homarr tRPC procedures safely through MCP. Use when adding or changing tRPC procedures under packages/api, adding MCP metadata, registering an eager MCP router, shaping tool inputs and descriptions, or reviewing MCP authorization, sensitive data, and destructive actions.
---
# MCP Integration
Expose a procedure only when an AI client should call it. Preserve the procedure's existing authorization and add MCP-specific safeguards for sensitive or destructive behavior.
## Assess exposure
Prefer:
- Data queries: list, search, get, health, stats, and summaries.
- User-requested actions with explicit permission checks and bounded inputs.
- Serializable results that do not reveal credentials or internal-only state.
Skip subscriptions, session/onboarding internals, file streams, blobs, credential values, and procedures whose authorization or audit behavior is unclear. Ask for a security review when a mutation is destructive, tenant boundaries are ambiguous, or returned data is sensitive.
## Add metadata
Place `.meta()` before `.input()` and `.query()` or `.mutation()`:
```typescript
getAll: protectedProcedure
.meta({
mcp:{
enabled: true,
description:"List resources the current user can access, including stable IDs used by resource_get",
No-input procedures can omit `.input()`; the MCP extractor accepts an empty object for them.
For mutations, state the effect and required permission in the description:
```typescript
remove: permissionRequiredProcedure
.meta({
mcp:{
enabled: true,
description:"Delete one resource by ID. Requires full access; get the ID from resource_all",
},
})
.input(z.object({id: z.string()}))
.mutation(async({ctx,input})=>{
// Perform the permission-scoped action.
});
```
## Write discoverable descriptions
Include every item that applies:
- What the tool returns or changes.
- Supported integrations or services.
- Required IDs and the tool that returns them.
- Permission requirements and the meaning of permission fields.
- Important bounds, confirmation semantics, or irreversible effects.
Use precise domain language. A description such as `Get calendar events` is insufficient because it omits supported services, required IDs, and result scope.
## Shape inputs for clients
- Use a top-level `z.object({...})` for parameterized tools.
- Bound strings, arrays, numeric ranges, pagination, and payload sizes.
- Use defaults where a safe, unsurprising value exists.
- Prefer explicit fields over a top-level union or discriminated union.
- Keep identifiers stable and describe how to obtain them.
- Keep secrets out of inputs unless the procedure is explicitly a credential-configuration flow with server-owned encryption and non-return guarantees.
## Register the eager router
The application router in `packages/api/src/root.ts` is lazy. MCP tool extraction is synchronous, so add an eager import and registration in `packages/api/src/mcp.ts`:
```typescript
import{resourceRouter}from"./router/resource";
exportconstmcpRouter=createTRPCRouter({
resource: resourceRouter,
});
```
Register the smallest router that owns the enabled procedures. If a parent router uses `lazy()`, import the needed subrouter directly. Metadata on a procedure is not enough; a procedure absent from `mcpRouter` is not exposed.
## Validate the tool surface
1. Confirm the procedure still enforces its ordinary auth and permission boundary.
2. Confirm `packages/api/src/mcp.ts` eagerly includes the procedure.
3. Confirm the generated tool name, description, and JSON schema are unambiguous.
4. When validating tool-list or extraction changes, use the focused API MCP spec: `bun run test packages/api/src/test/mcp.spec.ts`.
5. When validating protocol or transport changes, use the focused route spec: `bun run test apps/nextjs/src/app/api/mcp/[transport]/route.spec.ts`.
6. If manually probing `/api/mcp/<transport>`, use a scoped API key and avoid printing credentials or secret-bearing results.
description:Report a problem with a Custom JSX widget, its requests, preview, or editor
title:"bug: custom widget "
labels:["needs triage"]
body:
- type:markdown
attributes:
value:|
Share a small reproduction so we can distinguish widget code, upstream API, and Homarr problems.
Remove API keys, passwords, tokens, cookies, authorization headers, private URLs, and personal data from everything you share, including exports and screenshots. Replace private hosts with `https://example.invalid` while preserving relevant paths and response structure.
- type:textarea
id:behavior
attributes:
label:Actual and expected behavior
description:Describe what happens and what you expected. Include the exact error text and a sanitized screenshot if useful.
validations:
required:true
- type:textarea
id:reproduction
attributes:
label:Steps to reproduce
description:Include the affected request or control, relevant widget options, and whether this occurs in Test, preview, or on the dashboard. For layout issues, include the widget size and desktop or mobile viewport.
validations:
required:true
- type:textarea
id:definition
attributes:
label:Sanitized widget definition
description:Paste or attach the exported JSON with sources, requests, options, and JSX template. Prefer the smallest definition that reproduces the problem. If export fails, include the relevant JSX and request configuration. Check for private URLs and sensitive data before sharing.
validations:
required:true
- type:textarea
id:environment
attributes:
label:Environment
description:Include the exact Homarr release, commit SHA, or image digest; installation method; browser/version; and desktop or mobile OS. For request issues, include the source type (HTTP or integration), integration kind, and upstream service version. Do not include server URLs or credentials.
placeholder:|
Homarr version / commit / image digest:
Installation method:
Browser and OS:
Source type, integration kind, and service version:
validations:
required:true
- type:textarea
id:diagnostics
attributes:
label:Sanitized API response and logs
description:For request or data issues, include the request ID, method, relative API path, HTTP status, and a small response sample preserving field names and types. Add relevant Homarr or browser console errors. Remove sensitive values and private URLs; leave blank for purely visual issues.
Install [Act](https://nektosact.com/installation/index.html) and ensure Docker is running, then use:
```bash
bun run ci:act:fast
bun run ci:act:docker
bun run ci:act:all
bun run ci:act:cleanup
```
`fast` runs the database checks and quality gate and requires an `origin/dev` tracking ref. `docker` and `all` use the manual validation event to run those checks, build the container once, and run the converter boot/restart smoke test. `cleanup` validates the manual cleanup path without calling GitHub. The wrapper supports regular clones and Git worktrees, including non-default Docker contexts. Linux dependencies are kept in Docker volumes instead of overwriting host `node_modules`. The first run downloads the Act runner and Playwright browser dependencies.
For rootless Docker, set `ACT_TESTCONTAINERS_HOST` to a host address reachable from the runner when loopback callbacks are unavailable.
Remote caching is optional locally. Copy `.github/act/.secrets.example` to `.github/act/.secrets`, populate it with rotated credentials, and run with:
```bash
ACT_SECRET_FILE=.github/act/.secrets bun run ci:act:all
```
Act never publishes or deletes GHCR images. The secret file is ignored by Git.
## PR preview images
Successful non-draft pull requests from this repository publish an amd64 image:
The tag is replaced when the pull request is updated and deleted when it closes. Production Homarr images remain the supported multi-architecture releases. The `homarr-test` package must be configured as public in GHCR for unauthenticated testers.
## Dokploy preview cache
Keep `needs-demo` as the Dokploy preview label. Under the application's **Preview Build-time Secrets**, configure `TURBO_API`, `TURBO_TEAM`, `TURBO_TOKEN`, and `TURBO_REMOTE_CACHE_SIGNATURE_KEY`. Dokploy passes these to the Dockerfile as BuildKit secrets. Do not configure the token or signature key as Docker build arguments.
Rotate any credentials that have appeared in chat, logs, or committed configuration before enabling the cache.
if:github.actor_id == 158783068 || github.actor_id == 190541745 || github.actor_id == 210161987# Id of renovate bot and crowdin bot see https://api.github.com/users/homarr-renovate%5Bbot%5D and https://api.github.com/users/homarr-crowdin%5Bbot%5D and https://api.github.com/users/homarr-update-contributors%5Bbot%5D
--body "Detected change to be ${{ steps.semver.outputs.bump }}<br/>Bump version from ${{ steps.semver.outputs.current }} to ${{ steps.semver.outputs.next }}-beta" \
args:"Created a release PR ${{ steps.create-pull-request.outputs.url }} for version ${{ steps.semver.outputs.next }}-beta (new behaviour: ${{ steps.semver.outputs.bump }})"
args:"The release PR must be manually merged because the next version is a major version: ${{ steps.create-pull-request.outputs.url }} for release ${{ steps.semver.outputs.next }}-beta"
if git log --format='%s%n%b' "${latest_tag}..HEAD" \
| grep --extended-regexp --quiet '(^[[:alnum:]-]+(\([^)]*\))?!:)|(^BREAKING[ -]CHANGE:)'; then
exit 0
fi
echo "::error::package.json declares ${package_version}, but commits since ${latest_tag} contain no breaking-change signal. Use a conventional ! commit or BREAKING CHANGE footer so Semantic Release creates the intended major version."
# We read the last tag either from the created release or from the current branch, this is to rerun the deployment job for the currently released version when it failed
run:echo "No release-worthy commits were found since ${{ steps.semver.outputs.current }}." >> "$GITHUB_STEP_SUMMARY"
- name:Create pull request
run:'gh pr create --title "chore(release): automatic release ${{ steps.semver.outputs.next }}" --body "**This is an automatic release**.<br/>Manual action may be required for major bumps.<br/>Detected change to be ``${{ steps.semver.outputs.bump }}``<br/>Bump version from ``${{ steps.semver.outputs.current }}`` to ``${{ steps.semver.outputs.next }}``" --base main --head dev --label automerge'
--body "**This is an automatic release**.<br/>Manual action may be required for major bumps.<br/>Detected change to be ``${{ steps.semver.outputs.bump }}``<br/>Bump version from ``${{ steps.semver.outputs.current }}`` to ``${{ steps.semver.outputs.next }}``" \
args:"Created a release PR ${{ steps.create-pull-request.outputs.url }} for version ${{ steps.semver.outputs.next }} (new behaviour: ${{ steps.semver.outputs.bump }})"
args:"The release PR must be manually merged because the next version is a major version: ${{ steps.create-pull-request.outputs.url }} for release ${{ steps.semver.outputs.next }}"
│ ├── onboarding/ # Onboarding studio and setup flow
│ ├── workshop/ # Homarr-side Workshop client and contracts
│ └── cli/ # Node CLI for ops (brocli)
├── tooling/
│ ├── typescript/ # Base tsconfig
│ └── github/ # CI setup action
├── tools/
│ └── homarr-dev/ # Go CLI for local and PR Docker images
├── development/ # Dev docker-compose (Redis, MySQL, PostgreSQL)
├── e2e/ # E2E test specs
└── Dockerfile # Multi-stage production build
```
## Documentation
Add documentation only when users need the information to set up, use, or troubleshoot Homarr and cannot reasonably discover it from the UI or generated API schema. If a user does not need to know it, do not document it. A code, UI, or API change alone is never a reason to add docs.
Keep warranted documentation concise. Omit UI walkthroughs, visible control descriptions, internal implementation details, and anything already clear from the interface or schema.
Only after this user-need test passes, use these locations:
- New integration → `apps/docs/docs/integrations/<slug>/index.mdx` + `index.ts`
- New widget → `apps/docs/docs/widgets/<slug>/index.mdx` + `index.ts`
- Changed API → `apps/docs/docs/management/api/index.mdx`
- New/changed env vars → `apps/docs/docs/advanced/`
- New CLI commands → `apps/docs/docs/advanced/command-line/`
- New cron job → `apps/docs/docs/management/tasks.mdx`
## Monorepo Commands
-`bun run dev` — Next.js app only
-`bun run dev:cli -- dev` — run the developer CLI without installing a global binary
-`bun run db:seed` — seed default database data explicitly
-`bun run docker:dev:up` — start the Redis development service in the background
-`bun run dev:docs` — Fumadocs site only
-`bun run turbo build` — build all packages
-`bun run turbo build --filter=@homarr/docs` — build docs only
-`bun run turbo typecheck` — typecheck all packages
-`bun run lint` / `bun run format` — oxlint / oxfmt
## Code Style
- Lint: oxlint (not ESLint)
- Format: oxfmt (not Prettier)
- UI: Mantine (not Tailwind) — Tailwind is only used in docs app
- Mantine: use the `mantine` MCP server in `.mcp.json` for current v9 APIs before writing component code. Prefer built-in primitives (`Combobox`/`useCombobox`, the polymorphic `component` prop, `@mantine/hooks`) and check `packages/ui/` for existing conventions first.
- Icons: @tabler/icons-react
- Docs app can import from `@homarr/definitions` for shared types
- Run `bun run dev:cli -- dev` to browse local `homarr:*` images and remote PR images.
- Run `bun run dev:cli -- build <name>` from a Homarr checkout to build `homarr:<name>` with rebuild provenance.
- Run `bun run dev:cli -- build --pr <number>` to build a PR locally from a temporary checkout.
## Testing
Add tests only when requested, and only when the test is extremely relevant and will plausibly prevent a future regression in user-visible behavior or a security boundary. Do not add tests that restate implementation details or duplicate stronger coverage. For focused changes, run only the relevant existing or newly requested checks when validation is needed. Do not run broad test, Docker, or E2E suites by default.
## MCP servers
`.mcp.json` declares project-scoped MCP servers. Keep only keyless definitions there; credentials belong in local dotfiles. See [Mantine LLM tooling](https://mantine.dev/guides/llms/) for client-specific setup.
## Agent Skills
Portable skills live in `.agents/skills/`. Read the relevant `SKILL.md` before working in that domain; detailed references are loaded only when needed. Claude-compatible discovery is provided through `.claude/skills`.
-`codebase-context` — architecture, package boundaries, and shared utilities
-`documentation-sync` — user-critical documentation only
Turborepo configuration, task behavior, and CLI commands can vary between installed versions and may differ from your training data. Resolve the `turbo` package from this file's directory or relevant workspace; in monorepos, it may not be visible from the repository root. For example, run `node -p "require.resolve('turbo/package.json')"` from a workspace that depends on `turbo`.
Read `docs/README.md` inside that installed package first, then read the relevant pages from its `docs/` directory before changing Turborepo configuration or commands. Heed deprecation notices. These bundled docs match the installed package version and are available without network access.
This block is written and re-added by `turbo` before repository-scoped commands when an AI agent is detected. In the Turborepo source repository, its template is defined in `crates/turborepo-cli/src/cli/agent_guidance.rs`. Removing the managed block while updates are enabled means a later qualifying invocation will add it again. Set `"agentGuidance": false` in the root `turbo.json` or `turbo.jsonc` to opt out; this does not remove an existing block. When agent guidance is enabled, keep the block committed to avoid an uncommitted change on the next agent invocation. This repository disables agent guidance in `turbo.json`.
Self-hosted Homarr users who need to evaluate the project, install it, connect services, find extensions, and reference its APIs. They move between the homepage, documentation, Workshop, API reference, and project updates as one continuous experience.
## Product Purpose
The site explains what Homarr does and gives users reliable paths to install, configure, extend, and operate it. Success means users can identify where they are, move between major surfaces without relearning the navigation, and read technical content comfortably.
## Brand Personality
Practical, technical, welcoming.
## Anti-references
Avoid disconnected microsites, marketing-heavy copy, ornamental dashboards, hidden navigation, and dense long-form pages without visible structure.
## Design Principles
- Keep one recognizable Homarr identity across every surface.
- Let navigation topology match the task while keeping labels and behavior consistent.
- Prioritize readable line lengths, strong heading hierarchy, and visible document structure.
- Make primary paths obvious without adding decorative UI.
- Preserve useful functionality on mobile, keyboard navigation, and reduced-motion settings.
This version has breaking changes — APIs, conventions, and file structure may all differ from your training data. Read the relevant guide in `node_modules/next/dist/docs/` (resolved from this file's directory; in monorepos the `next` package may not be visible from the repo root) before writing any code. Heed deprecation notices.
This block is written and re-added by `next dev` — verify at `node_modules/next/dist/server/lib/generate-agent-files.js`. Removing it from a diff only re-creates the uncommitted change; committing it with your work keeps the tree clean.
<!-- END:nextjs-agent-rules -->
## Fumadocs workflow
- Start with `README.md`, `src/lib/source.ts`, `source.config.ts`, and the matching route under `src/app/`.
- Read current [Fumadocs documentation](https://fumadocs.dev/docs) with Context7 before changing framework APIs. Check the installed version and types; latest examples may use a different search adapter.
- Preserve static export: search uses `staticGET` with the matching static client. Do not introduce a request-time search server into this app.
- Keep integration/widget metadata in the existing typed definitions. Check both rendered HTML and processed Markdown when changing MDX components; JSX alone does not guarantee readable search or AI output.
- Run the README validation commands after content or navigation changes. Search results and table-of-contents links must resolve to rendered anchors.
- Kapa is an independent hosted assistant. Keep its keyboard shortcut disabled so it does not override docs search; follow the README for its build configuration and crawler setup.
People who run Homarr, administrators configuring self-hosted infrastructure, and contributors extending Homarr. They arrive with a concrete task such as installing Homarr, connecting a service, configuring a widget, managing users, or building a custom widget, and need an answer they can scan and apply quickly.
## Product Purpose
The documentation is the complete, authoritative operating guide for Homarr. It keeps setup steps, configuration contracts, screenshots, generated integration and widget metadata, API references, and Workshop tooling next to the application code. Success means a reader can find the right page quickly, understand the next action, copy reliable instructions, and verify the result without searching elsewhere.
## Brand Personality
Focused, dependable, and approachable. The writing is direct and practical. The interface should feel calm and familiar during long technical sessions while retaining Homarr's red identity.
## Anti-references
- Generic marketing layouts that replace Homarr's product identity with unrelated slogans or decorative conversion copy.
- Decorative dashboard aesthetics that compete with reading, including glass effects, large radii, card grids, and gratuitous motion.
- Generic generated-docs styling that obscures Homarr's identity or makes navigation feel detached from the main product.
- Dense reference portals that expose structure without giving readers a clear path through a task.
## Design Principles
1. Put the reader's task first: search, navigation, page outline, and the answer must remain easy to reach.
2. Preserve one source of truth: typed Homarr metadata and existing custom components stay connected to the monorepo.
3. Prefer familiar documentation patterns: the interface should disappear while a reader scans, copies, and follows instructions.
4. Make complete content feel manageable through strong hierarchy, progressive disclosure, and precise cross-links.
5. Treat machine readers as users through canonical Markdown, structured metadata, and standard LLM discovery endpoints.
6. Keep the root landing page recognizably Homarr: preserve its self-hosted dashboard promise and real product evidence while documentation routes remain task-first.
## Accessibility & Inclusion
Meet WCAG 2.2 AA for color contrast, keyboard operation, focus visibility, landmarks, headings, and interactive controls. Respect reduced-motion and system color-scheme preferences. Diagrams, screenshots, icons, and emoji must have meaningful text alternatives when they carry information.
[MDX](https://github.com/fuma-nama/fumadocs/blob/dev/packages/mdx/CHANGELOG.md), and
[OpenAPI](https://github.com/fuma-nama/fumadocs/blob/dev/packages/openapi/CHANGELOG.md) changelogs when upgrading.
- Core/UI and MDX now declare side effects for better tree shaking. UI also fixes tabs reverting to the URL hash.
- Core's `llms({ renderPage })` API provides the shared `index()`, `page()`, and `full()` exports.
- OpenAPI 12 replaces several custom renderer options and hooks. This site uses `createOpenAPIPage()` and
`getOpenAPIPageProps()` without those removed overrides; the built-in renderer retains all request languages.
The Scalar adapter receives an equivalent nonrecursive union for unconstrained JSON values to avoid recursive
schema expansion during rendering. The generated public OpenAPI document remains unchanged.
- MDX 15.4.3 fixes duplicate compilation output when its experimental build cache is enabled. We leave that cache
disabled: its current key checks source content without accounting for all compiler configuration changes.
- The existing macro collections, processed-Markdown component adapters, and static search remain supported.
Typed version roots and server-side MCP tools are available upstream but are not needed by this static site.
## Kapa AI
The compact Homarr mascot in the bottom-right opens **Ask AI**. It slowly spins when idle and gives a subtle claw movement and wink on hover. Animation respects reduced motion.
The launcher uses Homarr's existing public Kapa Website ID. The root layout loads the widget once across
client-side navigation and follows the docs light or dark theme. **Search** and `Ctrl+K` / `Cmd+K` always use Fumadocs search, including when Kapa is unavailable.
Set `KAPA_WEBSITE_ID` at build time to use a different Kapa Website Widget integration. An explicitly empty value
disables the widget, for example `KAPA_WEBSITE_ID= bun run --filter @homarr/docs build`. This is a public integration ID
included in exported pages, not a Kapa API key. Rebuild the static export after changing it.
In the Kapa dashboard, confirm the integration is live and enable the production and preview domains. Restoring the
script does not verify that the historical integration is still active. Follow the
[Kapa widget setup](https://docs.kapa.ai/integrations/website-widget/quickstart) for domain restrictions and CSP settings.
After changing the documentation structure, review the Website Crawl source in Kapa: use the deployed `/docs` URL as
the start URL and preview `main` as the content selector. Replace any old Docusaurus-specific selector and check that
titles, headings, code samples, and integration details appear in the extracted Markdown. Docs are exported as HTML;
the site also publishes `/sitemap.xml` and permits crawling in `/robots.txt`. The `/llms.txt` exports do not configure
Kapa ingestion automatically. See [Kapa source setup](https://docs.kapa.ai/getting-started/index-your-first-source).
Before publishing, check that Ask AI opens after navigating between pages, ask a documentation question, and verify
its citations point to the current docs. In a separate check, block `widget.kapa.ai` and confirm search still works.
## Release verification
The production image compiles the docs itself; it does not depend on a host `out/` directory.
but do not rewrite already-exported canonical URLs.
PostHog records SPA pageviews, installation clicks, searches with no results (including the search query), and Workshop
content downloads and submission starts/completions through `hog.homarr.dev`. Downloads include JSON files and
successful JSON/CSS clipboard copies. All events carry `site` and `source_path`. Autocapture, page-leave events, and
session replay are disabled; tracked URL query strings and fragments are removed. Localhost and `?analytics_test`
traffic carries `verification=true`; exclude it from production reports.
Carbon loads one visible placement after the desktop TOC, below API examples, or below content on other layouts.
The homepage is excluded at every viewport size. Navigation reloads the ad script; resizing changes placement only when crossing its breakpoint. Ad blockers
or no-fill responses must not prevent content, navigation, or search from working.
To preview Carbon locally, start docs with `NEXT_PUBLIC_ENABLE_CARBON_ADS=true bun run dev:docs`. Production builds
enable ads automatically. When the TOC collapses into a toggle, the ad appears below the content before “Edit this page”.
We are happy to announce that the documentation of Homarr has been migrated to Docusaurus - An opensource documentation tool.
The documentation has been re-written partly and includes now many animations, guides and additional crucial information.
We are still figuring things out and highly depend on your feedback.
{/* Historical excerpt boundary. */}
## Why we migrated
The default GitHub Wiki feature is decent - but very confusing to read.
Our possibilites for searching the documentation in the Wiki are highly limited.
Additionally, Users are unable to contribute to the documentation if they have not sufficient permissions.
We could not review any chanegs made to the documentation. This is why we migrated to Docusaurus.
## How you can contribute
The documentation will soon be merged into the master branch of Homarr.
You'll find a ``doc/`` directory in the root of Homarr.
We'll obviously not include this in our Docker images.
In the future, everyone may modify the documentation by creating a pull request.
After we've reviewed your changes, we'll merge your PR and this documentation will automatically update.
## TL;DR
The documentation on GitHub Wiki will soon disappear. It will be replaced with this Docusaurus-powered website. You'll find much more information here with higher detail and searchability.
We've been working hard to offer Homarr in many languages. The next update of Homarr will be using [i18next](https://www.i18next.com/) to translate almost any part of your Dashboard.
### We need your help
 <br/>
Yes, we need **YOU**! We want to make Homarr accessible to as many people as possible. Thus, we need to translate all content into many different languages. We need your help for that! [You can help us now to translate Homarr into as many languages as possible](/docs/community/translations).
{/* Historical excerpt boundary. */}
### About global translations / translations per user
We are aware that many people would like to configure translations globally or configure language per user. Currently, translations are stored in the Cookies, which will expire after some time (currently, it's set to 30 days). We will implement more granular language settings in future versions of Homarr, since it requires major changes to Homarr. Stay tuned! 😎
We'll continute to improve this system according to your feedback.
## Modules have been overhauled and renamed to widgets
Widgets are now much snappier and have been completely recoded.
This fixes some of the bugs a few users have previously experienced and improves overall customization.

## Fancy icon picker for speeding up your creation process
We've also added a new fancy icon picker, you can use to get access to any icons much faster.
We will add more icon sources in the future, as we made this icon picker very modular, and we plan to make local icons accessible via this picker very soon.
**It's finally here!** You can decorate your dashboard now with as many widgets as you want.
This applies to all widgets, but in the case of the IFrame widget, it enables you to:
- Integrate as many websites as you want
- Integrate monitoring tools like Grafana, Splunk or Elasticsearch
- Integrate your services (for example Sonarr, Radarr and Overseerr) directly into Homarr
## Edit mode password
We've added a setting to disable editing completely. This will allow you to restrict edit access without password-protection.
use the `EDIT_MODE_PASSWORD` ENV variable to set a password for edit mode.
turn on `DISABLE_EDIT_MODE=TRUE` to disable edit mode on startup.
{/* Historical excerpt boundary. */}
## 😎 Blazingly fast
We've implemented a new caching system, that will cache all data in your browser, meaning that Homarr will load within milliseconds, as if it was *powered by rust and blazingly fast* !!!
## Media request widgets (Overseerr & Jellyseerr)
We've implemented a new widget, that displays a history of all requests.
You can't approve from the widget yet, but we plan to add this in the future.
Since some have lost their changes, by refreshing the page in the edit mode. We've added a popup that will ask you to confirm you want to reload the page :
- Added mounting point ``/data`` for your database
- Require at least one administrator account on your instance. An onboarding process will automatically set this up.
- We've added analytics to the documentation and Homarr itself. We use [Umami Analytics](https://umami.is/) which collects anonymous data exclusively and protects your personal data rights. You can disable this using the environment variable ``DISABLE_ANALYTICS``.
One of the most requested features ever has been ___authentication___. We've completely *removed* the ``PASSWORD``, ``DISABLE_EDIT_MODE`` and ``EDIT_MODE_PASSWORD`` and replaced it by [a secure authentication system](https://next-auth.js.org/) that enables you to manage multiple users easily. We will allow for more detailed permission and user management soon. LDAP and OIDC integrations are currently also in progress and will arrive in a future version.
title: Homarr 1.0 - complete rewrite of all systems
date: 2024-09-23
authors:
- manuel-rw
---
We proudly present to you - Homarr ``1.0``.
We have started work on this new version back in 2023 and have worked tirelessly for months on this brand-new dashboard.
This isn't our usual update. ``1.0`` doesn't use the same code from previous versions, but has been written from the ground up.
Therefore, this version contains an unusual amount of breaking changes.
This will be a one time occurrence and version > ``1.0`` will have less breaking changes.
This time more important than every; **read the breaking changes** list and **create a backup** before updating.
Some third party installation methods may take multiple days or weeks to update to the new version.
## Changes at a glance
- **Asynchronous integrations system**: In versions prior to ``1.0``, we fetched data upon each request synchronously.
This means that the user had to wait for the requests to finish so any data would appear.
A page reload will trigger another request - and worse, this will happen for every user accessing your dashboard.
In ``1.0``, we have implemented a new system that gets rid of this system and scales linearly.
It will fetch data in background, even if you are not on the page.
- **Permissions system**: You can now use detailed permissions to allow or prohibit users from certain actions.
For easier management, you can organize users into groups and manage them all at the same time.
- **Integration testing**: Starting from ``1.0``, we will require you to test the integration before you can use it.
This guarantees that you've configured your integration correctly and helps you find errors faster.
- **Re-imagined widgets**: All widgets have been rewritten & improved. They offer much better performance,
reliability and scalability compared to < ``1.0``.
- **Improved performance**: We have upgraded the underlying framework of Homarr and take advantage
of an in-memory message bus and websockets to deliver data near realtime to you.
{/* Historical excerpt boundary. */}
## Breaking changes
<Alert type="warning">
Please consult the breaking changes list & upgrade guide if you used third-party installation methods.
</Alert>
- Docker Image
- Removed support for ``arm/v7`` -> upgrade to newer architectures as an alternative
- Renamed image from ``ghcr.io/ajnart/homarr`` to ``ghcr.io/homarr-labs/homarr``. We now use semantic versioning for our release numbers.
- Environment Variables
- New required env variable ``SECRET_ENCRYPTION_KEY`` used for encryption of secrets in the database
- Removed ``DISABLE_ANALYTICS``, can now be configured via UI
- Renamed ``AUTH_PROVIDER`` to ``AUTH_PROVIDERS``
- Renamed ``DATABASE_URL`` to ``DB_URL`` and made it optional conditionally. If ``DB_HOST`` is set, this is optional - otherwise it's required.
- Changed ``DOCKER_HOST`` to two new variables for hostnames (``DOCKER_HOSTNAMES``) and ports (``DOCKER_PORTS``). This allows for multiple docker integrations to be configured.
- ``AUTH_LOGOUT_REDIRECT_URL`` must now be a valid Url
- Increased minimum length of ``AUTH_OIDC_CLIENT_ID`` to 1
- Increased minimum length of ``AUTH_OIDC_CLIENT_SECRET`` to 1
- Removed ``AUTH_OIDC_TIMEOUT``
- Removed ``AUTH_LDAP_ADMIN_GROUP``, ``AUTH_LDAP_OWNER_GROUP``, ``AUTH_OIDC_ADMIN_GROUP`` and ``AUTH_OIDC_OWNER_GROUP`` -> groups are now managed without manual env binding
- ``CI`` will now also trigger an env validation skip like ``SKIP_ENV_VALIDATION``
- Mounts
- Removed ``/app/data/configs`` -> JSON configs have been removed in favour of relation databases
- Removed ``/app/public/icons`` -> Images can not be uploaded via the [media management](/docs/management/media/) in the UI
- Moved ``/data`` to ``/appdata`` and restructured contents (not backwards compatible!) -> Create a backup before migration of the entire directory to copy data to new Homarr
- Widgets:
- Calendar:
- Removed option ``Use Sonarr v4 API``
- Removed option ``Use Sonarr v5 API``
- Removed option ``Font Size``
- Removed option ``Hide week days``
- Download speed: We have removed this widget in favour of the brand new download widget which will also display this data.
- Dash. widget: We have removed this widget in favour of the System Health Widget. It will also display all data and some more than the old widget.
- Database:
- Changed schema of the database to a completely new schema. They are not backwards compatible. Please export the data from old Homarr in the UI and import it in ``1.0``
- Removed all usage of JSON files to store data. Import JSON files in Homarr ``1.0`` to import your dashboards.
- Removed ``schemaVersion`` from JSON and use the drizzle migrations to keep track of schema versions and take care of migrations.
- Web Server:
- Added an internal Nginx proxy to forward requests to two servers inside the container for technical reasons.
This could impact edge cases with iframes, cookies and custom setups.
- Dependencies
- We now depend on [Redis](https://redis.io/) as a real time messaging bus
- Development (relevant for contributors):
- Migrated to [pnpm](https://pnpm.io/)
- Migrated to a mono-repository
- All source code has been migrated to the [homarr-labs GitHub organization](https://github.com/homarr-labs)
- Platform specific breaking changes:
- TrueNAS
- The TrueCharts app chart is permanently deprecated and no longer maintained / supported. Please migrate to Electric Eel or a different Kubernetes platform.
- Unraid
- The Homarr app must be reinstalled. The existing directories on your disk are not compatible with Homarr 1.0. Export your old data in < `` 1.0`` and install using different file system locations.
- Kubernetes
- The official Helm chart has moved from https://github.com/oben01/charts to https://github.com/homarr-labs/charts
Loaded 100 of 4156 files, more files were not shown because too many files have changed in this diff.
Show more
Reference in new issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.