Windsurf
storage
Blob Storage Corrupting Uploaded Files in Windsurf App
Files uploaded to blob storage through your Windsurf-generated app become corrupted when downloaded. PDFs are unreadable, images show garbled data, ZIP files fail to extract, and documents open as random characters. The upload appears to succeed, but the stored file is damaged.
This is a particularly insidious bug because uploads appear successful — no errors are thrown, the file shows up in storage, and the file size may even look correct. But when a user downloads and opens the file, it's broken.
The corruption typically affects binary files (images, PDFs, ZIPs) while text files may work fine, which is a strong clue that the issue is related to encoding or stream handling.
Error Messages You Might See
Error: Invalid or corrupted PDF
Failed to decode image
Error: End-of-central-directory signature not found (ZIP)
The file is damaged and could not be opened
UTF-8 decode error on binary data
Common Causes
- UTF-8 encoding applied to binary data — Cascade read the file as a UTF-8 string instead of a Buffer, destroying binary data during encoding conversion
- Base64 double-encoding — The file is base64-encoded before upload, but the storage client also encodes it, resulting in double-encoded data
- Wrong content-type on upload — The file is uploaded with content-type: text/plain or application/json instead of its actual MIME type
- Stream not piped correctly — The readable stream is consumed (read to string) before being passed to the storage upload, instead of being piped directly
- Multipart boundary parsing error — The multipart form parser mishandles the file boundaries, truncating or appending extra bytes
How to Fix It
- Always use Buffer for binary files — Never call toString() on file data. Use Buffer objects or ArrayBuffer throughout the upload pipeline
- Set the correct content-type — Detect the MIME type using the file-type library or the upload's mimetype field, and pass it to the storage upload call
- Avoid double-encoding — Check if your storage client already handles encoding. Don't base64-encode if the client expects a Buffer or stream
- Pipe streams directly — Instead of reading the entire file into memory, pipe the upload stream directly to the storage write stream
- Verify with checksums — Generate an MD5 hash before upload and after download to confirm the file is identical. If hashes differ, the corruption is in the upload path
- Test with binary files specifically — Upload a small PNG or PDF and download it immediately. Open it to verify integrity before shipping to production
Real developers can help you.
Jaime Orts-Caroff
I'm a Senior Android developer, open to work in various fields
Matthew Jordan
I've been working at a large software company named Kainos for 2 years, and mainly specialise in Platform Engineering. I regularly enjoy working on software products outside of work, and I'm a huge fan of game development using Unity. I personally enjoy Python & C# in my spare time, but I also specialise in multiple different platform-related technologies from my day job.
Mehdi Ben Haddou
- Founder of Chessigma (1M+ users) & many small projects
- ex Founding Engineer @Uplane (YC F25)
- ex Software Engineer @Amazon and @Booking.com
Basel Issmail
’m a Senior Full-Stack Developer and Tech Lead with experience designing and building scalable web platforms. I work across the full development lifecycle, from translating business requirements into technical architecture to delivering reliable production systems.
My work focuses on modern web technologies, including TypeScript, Angular, Node.js, and cloud-based architectures. I enjoy solving complex technical problems and helping teams turn product ideas and prototypes into working platforms that can grow and scale.
In addition to development, I often collaborate closely with product managers, business analysts, designers, and QA teams to ensure that solutions align with both technical and business goals.
I enjoy working with startups and product teams where I can contribute both as a hands-on engineer and as a technical partner in designing and delivering impactful software.
David Olverson
Solo dev shipping production apps with AI-assisted development. I specialize in rescuing broken Lovable/Bolt/Cursor builds and taking them to production. 10+ apps shipped including SaaS CRMs, gaming platforms, real estate tools, and Discord bots. Stack: Next.js 16, TypeScript, Tailwind CSS, FastAPI, PostgreSQL, Prisma. I use Claude Code with 50+ custom skills for rapid delivery. Average turnaround: 2-4 weeks from broken prototype to production.
Tejas Chokhawala
Full-stack engineer with 5 years experience building production web apps using React, Next.js and TypeScript.
Focused on performance, clean architecture and shipping fast.
Experienced with Supabase/Postgres backends, Stripe billing, and building AI-assisted developer tools.
Meïr Ankri
Full-stack developer specializing in React / Next.js / Node.js with 6+ years of experience. I've worked across various sectors including automotive (Reezocar/Société Générale), healthcare (Medical Link SaaS), and e-commerce (Glasman). I build web apps end-to-end, from architecture to production, with a focus on scalability, performance, and code quality. I also mentor junior developers and contribute to technical decisions and code reviews.
MFox
Full-stack professional senior engineer (15+years). Extensive experience in software development, qa, and IP networking.
Prakash Prajapati
I’m a Senior Python Developer specializing in building secure, scalable, and highly available systems. I work primarily with Python, Django, FastAPI, Docker, PostgreSQL, and modern AI tooling such as PydanticAI, focusing on clean architecture, strong design principles, and reliable DevOps practices. I enjoy solving complex engineering problems and designing systems that are maintainable, resilient, and built to scale.
PawelPloszaj
I'm fronted developer with 10+ years of experience with big projects. I have small backend background too
You don't need to be technical. Just describe what's wrong and a verified developer will handle the rest.
Get HelpFrequently Asked Questions
Why do text files upload fine but images get corrupted?
Text files survive UTF-8 encoding because they're already text. Binary files (images, PDFs, ZIPs) contain byte sequences that are invalid UTF-8, so encoding them as strings destroys the data. Always handle binary files as Buffers.
How can I tell if corruption happens during upload or download?
Check the file directly in your storage dashboard (S3 console, Supabase dashboard). Download it from there. If it's corrupt in storage, the upload is the problem. If it's fine in storage but corrupt when downloaded through your app, the download handler is the issue.