Jauge de contexte, espace disque sur partage agrégé, panneau Fichiers

Trois défauts sans rapport entre eux, tous visibles à l'écran.

Jauge de contexte bloquée à « 0 / N (0 %) » : le comptage exact vient de
usage.prompt_tokens (include_usage), qu'un llama-server récent a cessé de
renvoyer. On note désormais si l'usage est arrivé ; sinon la fin de tour
publie l'estimation qui pilote déjà la compaction — approximative, mais
jamais absente. Et la souscription envoie la valeur courante juste après
le rattrapage : une page rechargée affichait zéro sur une discussion
pourtant pleine, faute d'événement ctx_used dans la fenêtre rejouée.

Espace disque : « il ne reste que 11,9 Go » sur un partage Unraid qui en
a des centaines. /mnt/user est servi par shfs, un FUSE qui agrège
plusieurs disques — statfs y renvoie l'espace d'un seul, et ce chiffre
REFUSAIT l'installation du modèle. Les systèmes de fichiers qui
approximent (FUSE, overlay, NFS, CIFS, Ceph, Gluster) sont reconnus par
leur magie : leur chiffre reste affiché, précédé d'un « ~ », mais ne
bloque plus rien.

Panneau Fichiers : le panneau entier défilait, emportant vers le haut
l'en-tête, le fil d'Ariane et le pied (occupation disque, bouton .zip)
dès qu'il y avait quelques fichiers. Seule la liste défile maintenant ;
le pied reste en bas et l'état vide se centre.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

frontend/node_modules/fastq/index.d.ts

@@ -0,0 +1,59 @@
+declare function fastq<C, T = any, R = any>(context: C, worker: fastq.worker<C, T, R>, concurrency: number): fastq.queue<T, R>
+declare function fastq<C, T = any, R = any>(worker: fastq.worker<C, T, R>, concurrency: number): fastq.queue<T, R>
+
+declare namespace fastq {
+  type worker<C, T = any, R = any> = (this: C, task: T, cb: fastq.done<R>) => void
+  type asyncWorker<C, T = any, R = any> = (this: C, task: T) => Promise<R>
+  type done<R = any> = (err: Error | null, result?: R) => void
+  type errorHandler<T = any> = (err: Error, task: T) => void
+
+  interface queue<T = any, R = any> {
+    /** Add a task at the end of the queue. `done(err, result)` will be called when the task was processed. */
+    push(task: T, done?: done<R>): void
+    /** Add a task at the beginning of the queue. `done(err, result)` will be called when the task was processed. */
+    unshift(task: T, done?: done<R>): void
+    /** Pause the processing of tasks. Currently worked tasks are not stopped. */
+    pause(): any
+    /** Resume the processing of tasks. */
+    resume(): any
+    running(): number
+    /** Returns `false` if there are tasks being processed or waiting to be processed. `true` otherwise. */
+    idle(): boolean
+    /** Returns the number of tasks waiting to be processed (in the queue). */
+    length(): number
+    /** Returns all the tasks be processed (in the queue). Returns empty array when there are no tasks */
+    getQueue(): T[]
+    /** Removes all tasks waiting to be processed, and reset `drain` to an empty function. */
+    kill(): any
+    /** Same than `kill` but the `drain` function will be called before reset to empty. */
+    killAndDrain(): any
+    /** Removes all tasks waiting to be processed, calls each task's callback with an abort error (rejects promises for promise-based queues), and resets `drain` to an empty function. */
+    abort(): any
+    /** Set a global error handler. `handler(err, task)` will be called each time a task is completed, `err` will be not null if the task has thrown an error. */
+    error(handler: errorHandler<T>): void
+    /** Property that returns the number of concurrent tasks that could be executed in parallel. It can be altered at runtime. */
+    concurrency: number
+    /** Property (Read-Only) that returns `true` when the queue is in a paused state. */
+    readonly paused: boolean
+    /** Function that will be called when the last item from the queue has been processed by a worker. It can be altered at runtime. */
+    drain(): any
+    /** Function that will be called when the last item from the queue has been assigned to a worker. It can be altered at runtime. */
+    empty: () => void
+    /** Function that will be called when the queue hits the concurrency limit. It can be altered at runtime. */
+    saturated: () => void
+  }
+
+  interface queueAsPromised<T = any, R = any> extends queue<T, R> {
+    /** Add a task at the end of the queue. The returned `Promise` will be fulfilled (rejected) when the task is completed successfully (unsuccessfully). */
+    push(task: T): Promise<R>
+    /** Add a task at the beginning of the queue. The returned `Promise` will be fulfilled (rejected) when the task is completed successfully (unsuccessfully). */
+    unshift(task: T): Promise<R>
+    /** Wait for the queue to be drained. The returned `Promise` will be resolved when all tasks in the queue have been processed by a worker. */
+    drained(): Promise<void>
+  }
+
+  function promise<C, T = any, R = any>(context: C, worker: fastq.asyncWorker<C, T, R>, concurrency: number): fastq.queueAsPromised<T, R>
+  function promise<C, T = any, R = any>(worker: fastq.asyncWorker<C, T, R>, concurrency: number): fastq.queueAsPromised<T, R>
+}
+
+export = fastq