Bkend Storage logo

Bkend Storage

Organization
ww-w-ai
bkend-storage

bkend.ai file storage expert skill. Covers single/multiple/multipart file upload via Presigned URL, file download (CDN vs Presigned), 4 visibility levels (public/private/protected/shared), bucket management, and file metadata. Triggers: file upload, download, presigned, bucket, storage, CDN, image, 파일 업로드, 다운로드, 버킷, 스토리지, 이미지, ファイルアップロード, ダウンロード, バケット, ストレージ, 文件上传, 下载, 存储桶, 存储, carga de archivos, descarga, almacenamiento, cubo, telechargement, televersement, stockage, seau, Datei-Upload, Download, Speicher, Bucket, caricamento file, download, archiviazione, bucket Do NOT use for: authentication (use bkend-auth), database queries (use bkend-data), MCP setup (use bkend-mcp)

Overview

Publisherww-w-ai
Repositorybkit-gemini
Skill namebkend-storage
Stars
66
Forks
16
Bundled files
Instructions only
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by ww-w-ai on GitHub. Read the source before you install it.

Installation

Install the Bkend Storage AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/ww-w-ai/bkit-gemini.git /tmp/bkit-gemini
mkdir -p .claude/skills
cp -r /tmp/bkit-gemini/skills/bkend-storage .claude/skills/bkend-storage
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Bkend Storage in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Bkend Storage on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Bkend Storage is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

bkend-storage: File Storage Expert Skill

Complete file storage management for bkend.ai projects using S3-based Presigned URLs

1. Overview

bkend.ai provides a fully managed file storage service built on S3-compatible object storage with a Presigned URL upload pattern and CDN delivery for public files.

Key characteristics:

  • S3-based storage with Presigned URL upload pattern
  • CDN delivery for public file access
  • 4 visibility levels: public, private, protected, shared
  • Multipart upload support for large files
  • File metadata management via REST API

IMPORTANT: There are no MCP tools for storage operations. All file storage interactions use the REST API only.

2. Single File Upload (3-Step Pattern)

Every file upload follows a strict 3-step process: get a presigned URL, upload the file binary, then register the metadata.

Step 1: Request a Presigned URL

http
POST /v1/files/presigned-url
Content-Type: application/json

{
  "fileName": "profile-photo.jpg",
  "contentType": "image/jpeg",
  "visibility": "public",
  "tableName": "users",
  "recordId": "rec_abc123"
}
  • fileName (required): Original file name
  • contentType (required): MIME type of the file
  • visibility (required): One of public, private, protected, shared
  • tableName (optional): Associate file with a specific table
  • recordId (optional): Associate file with a specific record

Response:

json
{
  "success": true,
  "data": {
    "presignedUrl": "https://s3.amazonaws.com/bucket/...",
    "fileKey": "projects/proj_001/files/abc123/profile-photo.jpg"
  }
}

Step 2: Upload File Binary to Presigned URL

http
PUT {presignedUrl}
Content-Type: image/jpeg

<file binary data>

Upload the raw file binary directly to the presigned URL returned in Step 1. The Content-Type header must match the contentType from Step 1.

Step 3: Register File Metadata

http
POST /v1/files
Content-Type: application/json

{
  "fileKey": "projects/proj_001/files/abc123/profile-photo.jpg",
  "fileName": "profile-photo.jpg",
  "contentType": "image/jpeg",
  "size": 245678,
  "visibility": "public"
}

Response:

json
{
  "success": true,
  "data": {
    "id": "file_xyz789",
    "fileKey": "projects/proj_001/files/abc123/profile-photo.jpg",
    "fileName": "profile-photo.jpg",
    "contentType": "image/jpeg",
    "size": 245678,
    "visibility": "public",
    "url": "https://cdn.bkend.ai/projects/proj_001/files/abc123/profile-photo.jpg",
    "createdBy": "usr_abc",
    "createdAt": "2025-01-15T09:30:00Z"
  }
}

3. Multiple File Upload

For uploading multiple files, follow the same 3-step pattern for each file. Use the batch endpoint to request multiple presigned URLs at once:

http
POST /v1/files/presigned-urls
Content-Type: application/json

{
  "files": [
    {
      "fileName": "photo-1.jpg",
      "contentType": "image/jpeg",
      "visibility": "public"
    },
    {
      "fileName": "photo-2.png",
      "contentType": "image/png",
      "visibility": "public"
    },
    {
      "fileName": "document.pdf",
      "contentType": "application/pdf",
      "visibility": "private"
    }
  ]
}

Response:

json
{
  "success": true,
  "data": [
    {
      "fileName": "photo-1.jpg",
      "presignedUrl": "https://s3.amazonaws.com/...",
      "fileKey": "projects/proj_001/files/abc/photo-1.jpg"
    },
    {
      "fileName": "photo-2.png",
      "presignedUrl": "https://s3.amazonaws.com/...",
      "fileKey": "projects/proj_001/files/def/photo-2.png"
    },
    {
      "fileName": "document.pdf",
      "presignedUrl": "https://s3.amazonaws.com/...",
      "fileKey": "projects/proj_001/files/ghi/document.pdf"
    }
  ]
}

Then upload each file binary to its presigned URL (Step 2) and register each file's metadata (Step 3).

4. Multipart Upload (Large Files)

For large files, use multipart upload to split the file into smaller parts and upload them individually.

4.1 Initialize Multipart Upload

http
POST /v1/files/multipart/init
Content-Type: application/json

{
  "fileName": "large-video.mp4",
  "contentType": "video/mp4",
  "visibility": "private"
}

Response:

json
{
  "success": true,
  "data": {
    "uploadId": "upload_abc123",
    "fileKey": "projects/proj_001/files/xyz/large-video.mp4"
  }
}

4.2 Get Presigned URL for Each Part

http
POST /v1/files/multipart/presigned-url
Content-Type: application/json

{
  "uploadId": "upload_abc123",
  "fileKey": "projects/proj_001/files/xyz/large-video.mp4",
  "partNumber": 1
}

Response:

json
{
  "success": true,
  "data": {
    "presignedUrl": "https://s3.amazonaws.com/...",
    "partNumber": 1
  }
}

4.3 Upload Each Part

http
PUT {presignedUrl}
Content-Type: application/octet-stream

<part binary data>

The response includes an ETag header that must be saved for the completion step.

4.4 Complete Multipart Upload

After all parts are uploaded, finalize the upload:

http
POST /v1/files/multipart/complete
Content-Type: application/json

{
  "uploadId": "upload_abc123",
  "fileKey": "projects/proj_001/files/xyz/large-video.mp4",
  "parts": [
    { "partNumber": 1, "etag": "\"abc123\"" },
    { "partNumber": 2, "etag": "\"def456\"" },
    { "partNumber": 3, "etag": "\"ghi789\"" }
  ]
}

4.5 Abort Multipart Upload

Cancel an in-progress multipart upload:

http
POST /v1/files/multipart/abort
Content-Type: application/json

{
  "uploadId": "upload_abc123",
  "fileKey": "projects/proj_001/files/xyz/large-video.mp4"
}

5. File Download

5.1 CDN Download (Public Files)

Public files are served via CDN for fast global delivery:

GET https://cdn.bkend.ai/{fileKey}

Example:

GET https://cdn.bkend.ai/projects/proj_001/files/abc123/profile-photo.jpg

No authentication is required for public CDN URLs.

5.2 Presigned Download (Private Files)

Private, protected, and shared files require a presigned download URL:

http
GET /v1/files/{fileId}/download

Response:

json
{
  "success": true,
  "data": {
    "downloadUrl": "https://s3.amazonaws.com/...?X-Amz-Signature=...",
    "expiresIn": 3600
  }
}

The presigned download URL is temporary and expires after the specified expiresIn duration (in seconds).

6. Visibility Levels

bkend.ai supports 4 visibility levels that control who can access uploaded files:

VisibilityAccessCDN AvailableDescription
publicAnyoneYesAccessible to everyone via CDN URL. No authentication required.
privateOwner onlyNoOnly the file owner (uploader) can access via presigned download.
protectedAuthenticated usersNoAny authenticated user in the project can access via presigned download.
sharedSpecified usersNoOnly users explicitly granted access can download via presigned URL.

Choosing the right visibility:

  • public: Profile pictures, product images, public assets
  • private: Personal documents, private exports, user-specific files
  • protected: Internal team documents, shared project resources
  • shared: Files shared with specific collaborators

7. File Metadata Management

7.1 Get File Metadata

http
GET /v1/files/:fileId

Response:

json
{
  "success": true,
  "data": {
    "id": "file_xyz789",
    "fileKey": "projects/proj_001/files/abc123/profile-photo.jpg",
    "fileName": "profile-photo.jpg",
    "contentType": "image/jpeg",
    "size": 245678,
    "visibility": "public",
    "url": "https://cdn.bkend.ai/projects/proj_001/files/abc123/profile-photo.jpg",
    "tableName": "users",
    "recordId": "rec_abc123",
    "createdBy": "usr_abc",
    "createdAt": "2025-01-15T09:30:00Z",
    "updatedAt": "2025-01-15T09:30:00Z"
  }
}

7.2 Update File Metadata

http
PUT /v1/files/:fileId
Content-Type: application/json

{
  "fileName": "new-file-name.jpg",
  "visibility": "private"
}

7.3 Delete File

http
DELETE /v1/files/:fileId

Response:

json
{
  "success": true,
  "data": {
    "id": "file_xyz789",
    "deleted": true
  }
}

7.4 List Files

http
GET /v1/files?tableName=users&recordId=rec_abc123&visibility=public&limit=20&cursor=file_last_id

Query parameters:

ParameterTypeDescription
tableNamestringFilter by associated table
recordIdstringFilter by associated record
visibilitystringFilter by visibility level
limitintNumber of results (max 100)
cursorstringCursor for pagination

8. Frontend Upload Pattern

Use the 3-step pattern with bkendFetch in your frontend code:

typescript
// lib/storage.ts
import { bkendFetch } from '@/lib/bkend';

interface UploadOptions {
  fileName: string;
  contentType: string;
  visibility: 'public' | 'private' | 'protected' | 'shared';
  tableName?: string;
  recordId?: string;
}

export async function uploadFile(file: File, options: UploadOptions) {
  // Step 1: Get presigned URL
  const presignedRes = await bkendFetch('/v1/files/presigned-url', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
      fileName: options.fileName,
      contentType: options.contentType,
      visibility: options.visibility,
      tableName: options.tableName,
      recordId: options.recordId,
    }),
  });
  const { presignedUrl, fileKey } = (await presignedRes.json()).data;

  // Step 2: Upload file binary to presigned URL
  await fetch(presignedUrl, {
    method: 'PUT',
    headers: { 'Content-Type': options.contentType },
    body: file,
  });

  // Step 3: Register file metadata
  const registerRes = await bkendFetch('/v1/files', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
      fileKey,
      fileName: options.fileName,
      contentType: options.contentType,
      size: file.size,
      visibility: options.visibility,
    }),
  });
  const registeredFile = (await registerRes.json()).data;

  return registeredFile;
}

// Usage
const file = document.querySelector<HTMLInputElement>('#file-input')?.files?.[0];
if (file) {
  const result = await uploadFile(file, {
    fileName: file.name,
    contentType: file.type,
    visibility: 'public',
    tableName: 'users',
    recordId: 'rec_abc123',
  });
  console.log('Uploaded:', result.url);
}

9. Image Upload with Preview (React Component)

tsx
// components/ImageUpload.tsx
'use client';

import { useState, useCallback } from 'react';
import { uploadFile } from '@/lib/storage';

interface ImageUploadProps {
  onUpload: (fileData: { id: string; url: string; fileName: string }) => void;
  visibility?: 'public' | 'private' | 'protected' | 'shared';
  tableName?: string;
  recordId?: string;
}

export function ImageUpload({
  onUpload,
  visibility = 'public',
  tableName,
  recordId,
}: ImageUploadProps) {
  const [preview, setPreview] = useState<string | null>(null);
  const [uploading, setUploading] = useState(false);
  const [error, setError] = useState<string | null>(null);

  const handleFileSelect = useCallback(
    async (e: React.ChangeEvent<HTMLInputElement>) => {
      const file = e.target.files?.[0];
      if (!file) return;

      // Validate file type
      if (!file.type.startsWith('image/')) {
        setError('Please select an image file');
        return;
      }

      // Validate file size (max 10MB)
      if (file.size > 10 * 1024 * 1024) {
        setError('File size must be less than 10MB');
        return;
      }

      // Show preview
      const reader = new FileReader();
      reader.onload = (event) => {
        setPreview(event.target?.result as string);
      };
      reader.readAsDataURL(file);

      // Upload
      setUploading(true);
      setError(null);

      try {
        const result = await uploadFile(file, {
          fileName: file.name,
          contentType: file.type,
          visibility,
          tableName,
          recordId,
        });

        onUpload({
          id: result.id,
          url: result.url,
          fileName: result.fileName,
        });
      } catch (err) {
        setError('Upload failed. Please try again.');
        setPreview(null);
      } finally {
        setUploading(false);
      }
    },
    [visibility, tableName, recordId, onUpload]
  );

  return (
    <div className="image-upload">
      <label className="upload-label">
        <input
          type="file"
          accept="image/*"
          onChange={handleFileSelect}
          disabled={uploading}
          className="hidden"
        />
        {preview ? (
          <img src={preview} alt="Preview" className="upload-preview" />
        ) : (
          <div className="upload-placeholder">
            <span>Click to upload image</span>
          </div>
        )}
      </label>

      {uploading && <p className="upload-status">Uploading...</p>}
      {error && <p className="upload-error">{error}</p>}
    </div>
  );
}

10. Error Codes

Error CodeHTTP StatusDescription
FILE_TOO_LARGE413File exceeds the maximum allowed size
INVALID_FILE_TYPE400The file MIME type is not allowed
PRESIGNED_URL_EXPIRED400The presigned URL has expired before upload completed
FILE_NOT_FOUND404The specified file ID does not exist

Error response format:

json
{
  "success": false,
  "error": {
    "code": "FILE_TOO_LARGE",
    "message": "File size exceeds the maximum limit of 100MB",
    "details": {
      "maxSize": 104857600,
      "actualSize": 157286400
    }
  }
}

Quick Reference

Common Workflows

  1. Upload a single file -> 3-step: presigned URL, PUT binary, register metadata
  2. Upload multiple files -> Batch presigned URLs, then Step 2 & 3 for each
  3. Upload large files -> Multipart: init, get part URLs, upload parts, complete
  4. Access public files -> CDN URL: https://cdn.bkend.ai/{fileKey}
  5. Access private files -> Presigned download: GET /v1/files/{fileId}/download
  6. Update file visibility -> PUT /v1/files/:fileId with new visibility
  7. Delete a file -> DELETE /v1/files/:fileId
  8. List files for a record -> GET /v1/files?tableName=x&recordId=y

Frequently asked questions

What does the Bkend Storage AI skill do?

bkend.ai file storage expert skill. Covers single/multiple/multipart file upload via Presigned URL, file download (CDN vs Presigned), 4 visibility levels (public/private/protected/shared), bucket management, and file metadata. Triggers: file upload, download, presigned, bucket, storage, CDN, image, 파일 업로드, 다운로드, 버킷, 스토리지, 이미지, ファイルアップロード, ダウンロード, バケット, ストレージ, 文件上传, 下载, 存储桶, 存储, carga de archivos, descarga, almacenamiento, cubo, telechargement, televersement, stockage, seau, Datei-Upload, Download, Speicher, Bucket, caricamento file, download, archiviazione, bucket Do NOT use for: authentica...

Why use Bkend Storage on TypingMind?

Because you install it once and use it with any model. Bkend Storage is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Bkend Storage in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/ww-w-ai/bkit-gemini/tree/main/skills/bkend-storage. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Bkend Storage?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Bkend Storage?

As many as you like. As long as a model supports skills, you can use Bkend Storage with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Bkend Storage AI skill free?

Yes. It is published on GitHub by ww-w-ai under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇