I have a perfectly working workflow, which creates a markdown file, so a binary file as output.
Downloading this file is ok, and it displays a well formatted markdown
But … when I click on view, I get this error message ‘This content is blocked. Contact the site owner to fix the issue.’
I have the updated to the latest N8N version (cloud version) and still the same error. Incognito window, same error message
@peter_wett your file and workflow are fine, this is n8ns own doing. the binary View serves the file under a sandbox content-security-policy to isolate user-generated content, and the browser blocks rendering the markdown under that sandbox, which is the “content is blocked” message youre seeing. Download works because its sent as an attachment so it skips that sandboxed inline render. its baked into the view endpoint, not something you can toggle off on Cloud, so for a .md the download is the reliable path. nothing to fix on your end.
ok, I understand.
But why was it working until last week?
Sounds weird, for an automation, but I need the ‘view’ instead of the ‘download’
I copy and paste it somewhere else (a local file on my laptop). Ok, I know some ways to make it work, but too much hassle. I was happy with how it was last week.
@peter_wett yeah that lines up, a recent n8n update tightened the sandbox on the binary view and Cloud auto-updated you into it last week, so the same file that used to render now gets blocked. cant say if that was intentional hardening or an accidental regression, but since it broke a working flow its worth a quick bug report to n8n. for your copy-paste use though you dont actually need the View, keep the markdown as a normal text field on the item (its a string before you convert it to a file) and copy it straight from the output panel json/schema view, no binary, no sandbox.
done a test with another workflow and the ‘view’ does work when it comes to a .jpeg
I guess I will have to live it with and add a ugly, but working, solution after this block.
I love challenges
The “View” button tries to render the file in an iframe, and browsers block iframes for text/markdown MIME type due to CSP - that’s the root cause. The clean fix: add a “Convert Binary” node (set operation to “Binary to Text”, encoding UTF-8) right after your markdown file node. This gives you the raw markdown content as a string in $json.data. From there you can read it directly in the n8n output panel as text, or pass it to a subsequent node. You won’t get the visual preview, but you skip the browser block entirely.
