BrewPage

nice and useful product for me!

I built BrewPage because sharing an HTML report or an AI-generated page always meant a repo, a deploy or an account. Now it is one paste or one POST request, and agents can publish through MCP. Feedback on the API and MCP flow is very welcome.

One POST and a link back is a nice fit for CI test reports, where nobody needs the page after a week anyway. Does the API response include the expiry date? Then a CI job could print it next to the link, and nobody gets surprised when a shared report disappears.

The expiry split is the part I would want clearer: 30 days with no account versus 90 days with a free profile. If an agent publishes through the MCP server (npm brewpage-mcp or the remote /mcp endpoint) without a profile, does that page still expire at 30 days? And does the owner token still update the same URL only before expiry, or can a later GitHub Action overwrite a CI report link after it has lapsed?

The one-POST-request approach plus the MCP server is a neat way to let agents publish pages without accounts. It would be helpful to document the exact expiry and size limits near the API examples so developers can plan CI report hosting.

Really impressive product! I love how BrewPage makes publishing and sharing HTML pages, Markdown files and static websites so simple without requiring an account or complicated deployment setup. The REST API and MCP integration are especially useful for developers and AI-powered workflows. It's a practical tool that saves time and removes unnecessary friction. Definitely worth checking out. Great work by the maker!

Publishing an AI-generated HTML artifact as a link with one POST request is very handy. Could you offer an optional short custom slug, so shared pages are easier to recognise in a team chat?

An MCP server plus a plain POST API makes this handy for letting an agent share a preview. Since publications expire after 30 days without an account, does the response include the expiry date, so the agent can warn the user before the link dies?

BrewPage is a really clever solution to a common developer workflow problem. Being able to publish HTML, Markdown, static sites or AI-generated artifacts and get an instant shareable HTTPS link without creating an account removes a lot of unnecessary friction. The REST API and MCP integration are particularly interesting. Allowing Claude, ChatGPT and other AI agents to publish content directly opens up some great possibilities for automated workflows, CI reports and quick prototypes. I also like the ability to update an existing publication without changing its URL. One question: have you considered adding optional custom domains or longer-term hosting for projects that outgrow the temporary publishing model? Really practical tool with a lot of potential!

Great app. Just wanted to know what do you mean by AI artifacts? Can I also host claude ones? Is there any plugin for claude?

Since the owner token keeps the same URL, does updating a page reset its 30-day expiry, or does it still count from the first publish? And is there a size limit per site?

The owner token for updating at the same URL is a nice touch for CI reports, because the link posted in a PR comment can stay the same across reruns. Is there a size cap for a multi-file static site upload? And can the JSON/KV hosting hand out a read-only key, so an agent can share data without exposing the token that can overwrite it?

Looks great, very useful for me. Will give it a try!

The no-signup part is the real differentiator — most quick hosting tools still want an account before you can share a file. Pasting a static HTML export and getting an HTTPS link in seconds is exactly the right workflow for AI artifacts and one-off demos. A browser-generated key to list and revoke past links later would make it even safer to use without an account.

No-signup static hosting is a genuinely useful primitive - the friction drop from zero accounts is huge for quick shares and docs. The obvious tension is abuse: anonymous hosting attracts phishing pages and malware drops, and that eventually lands on your domain reputation and host. Worth thinking through upfront: rate limits per IP, a report path that's fast to act on, and maybe a content hash blocklist shared with other similar services. Cheap to add now, painful to retrofit once you're the go-to free host.

Really clean approach to zero-friction artifact hosting. The REST API endpoint for instant HTTPS previews is especially useful for agentic workflows and local CLI pipelines that need quick ephemeral share links. Are the KV stores persistent across cold starts or TTL-evicted?

The editor's generated password is four characters, the minimum the API accepts, and the agent guidance suggests a private namespace plus a password for secrets. Are wrong password attempts limited per page, or only by the per-IP read limit in the FAQ?

The MCP server is the interesting part for me. I run one for my own app, and the biggest lesson was that every write tool should return the full resulting state in structuredContent, not just "done". Otherwise the model guesses the rest and gets it wrong. For BrewPage that would mean returning the URL, the exact expiry timestamp and how to extend or delete it, so the agent can tell the user "this link dies on Dec 1" instead of inventing it. How are you handling abuse with no-signup hosting? Phishing pages tend to show up fast on open hosts.

The owner token that lets you update or delete a page at the same URL is handy for CI reports that get regenerated on every run. Does the GitHub Action reuse that token so each build overwrites the previous report instead of creating a new link?

Anonymous HTML hosting with scripts allowed is a natural target for phishing pages, so I am curious how uploads are isolated. Are user pages served from a separate origin than brewpage.app, so an uploaded script cannot reach the profile session or the owner-token UI? A short note on that in the docs would help anyone deciding whether to publish agent output here.

Splitting expiry into 30 days without an account and 90 days with a free profile makes the tradeoff easy to understand. The owner token for updating or deleting at the same URL is handy for fixing a shared report after it goes out.

Nice to see a hosting tool that skips signup entirely. For CI reports, it would help to show the exact expiry date in the API response so a pipeline can warn before a link lapses. Is there a per-file size limit for the multi-file static sites?

Useful product, hoping to make sure of it soon

Sharing a generated report as a plain HTTPS link is a real pain point; we ship PDF and HTML reports from a browser app and clients often just want a link they can open on a phone. How does the page behave when the HTML references external CSS or images that are not in the upload, for example a report that pulls a font from a CDN? I would also find it useful if the API response said whether a page is still within its expiry window, so a CI job could re-publish only when needed.

The one-POST-request flow is a nice touch for AI-generated reports. On the API feedback: returning the expiry date in the response (someone asked above) would make it easy for agents to warn users before a link dies. A small thought for shared pages: an option to add a noindex tag, so a temporary report doesn't end up in search results by accident.

Hosting HTML/Markdown/JSON with no sign-up is a great low-friction idea. I use the same "no account needed" approach for a free tool, and it really lowers the barrier. How do you handle abuse or spam without accounts?

Hosting HTML/Markdown/JSON with no sign-up is a great low-friction idea. I use the same "no account needed" approach for a free tool, and it really lowers the barrier. How do you handle abuse or spam without accounts?

Clean, no-signup hosting for static files is exactly what many dev workarounds need. Two ideas: add a simple URL shortener per file so teams can share stable links without a domain, and a per-path expiry option for temporary assets like drafts or test builds.

Getting a link from one POST is great for sharing price-comparison snapshots with clients. Can a page be marked noindex, or does it send an X-Robots-Tag header, so temporary reports stay out of Google before they expire?

The no-signup flow with a 30-day expiry (90 days with a free profile) makes BrewPage handy for sharing AI-generated HTML artifacts quickly. I like that the MCP server lets an agent publish on its own. Is there a rate limit or size cap for anonymous uploads, and can the owner token be rotated if it leaks?

For CI reports re-published through the owner token, how is caching handled? A short Cache-Control on index.html and a long one on assets would make the same link show the new run immediately instead of a stale copy from a CDN or the browser.

BrewPage has an interesting approach to making content publishing simple while also supporting API and MCP workflows. Turning HTML, Markdown, JSON, and AI-generated artifacts into shareable HTTPS links could be especially useful for developers and AI agents. I'm curious, what's the best way to manage published links for projects that need to stay accessible over a longer period?

When the owner token deletes a page, is it purged from the CDN right away, or can a cached copy keep serving until the TTL runs out? For people sharing something sensitive, like a client report or a draft agreement, instant deletion matters more to me than the 30/90-day expiry.

No-signup hosting is handy for sharing a quick HTML prototype with a client. How long do anonymous pages stay up, and can I update or delete a page later without an account (for example with an edit token)? A clear retention note on the landing page would make it easier to trust for demos.

The outlier score (views relative to a creator's own median) is a smarter signal than raw view counts for finding what really worked. How does it handle smaller niche accounts with only a handful of videos, where the median can be noisy?

Publishing JSON and key-value data with an expiry (30 days anonymous, 90 with a free profile) is a handy fit for agents that need to read small state back. The MCP server and GitHub Action make it easy to drop into a CI or Claude workflow without signup.

The one-POST-request approach plus the MCP server is a neat way to let agents publish pages without accounts. It would be helpful to document the exact expiry and size limits near the API examples so developers can plan CI report hosting.

The one-POST-request approach plus the MCP server is a neat way to let agents publish pages without accounts. It would be helpful to document the exact expiry and size limits near the API examples so developers can plan CI report hosting.

looks good. very useful tool

nice and useful product for me!

Free file hosting with no signup is the kind of zero-friction tool I reach for when I just need to publish a changelog, a JSON config or a quick Markdown page without touching a repo. Two things that would make this a primary instead of a backup: signed URLs with an expiry for anything that sho my own dashboard. Are the published paths predictable (e.g. your-slug.brewpage.app), and is t here an API key or token flow for CI pipelines that want to push updates without touching the site?uldn't be cached forever, and a webhook callback when a hosted file is fetched so I can hook a lightweight analytics page into

The 30/90-day expiry is the design choice I'd want to hear more about - it frames this as a link you hand someone rather than hosting, and it keeps the abuse surface small. Does an expired page return 410, or redirect to a notice explaining what happened? The second is friendlier for anyone who saved the link.

The no-signup HTTPS link for HTML and Markdown is exactly what I would want for sharing quick prototypes and reports with the team. The MCP server for agents is a thoughtful touch too. — Rob Eley, RFPG