Help Centre

Getting started with Rawtell, answers to common questions, and known limitations of the forensic JSON debugging tool. If you are here because a parse error brought you — the tool itself is at /app →

Getting started

Inspecting a URL

Navigate to the App page, select the Inspect URL tab, paste a full URL (including https://) into the input field, and click Inspect. Rawtell fetche the URL and returns:

Parsing JSON text

Select the Parse text tab, paste your JSON blob into the text area, and click Parse. The entire parse happens in your browser — the text is never sent to any server. Rawtell highlights the exact location of any parse error and explains what was expected there in plain English.

Frequently asked questions

My JSON parses fine in other tools — why does Rawtell say it is invalid?
Different tools use different JSON parsers. Some parsers (notably JavaScript's native JSON.parse in most browsers) are strict. Rawtell uses a strict parser and will reject:
  • Trailing commas in objects or arrays
  • Single-quoted strings ('value' instead of "value")
  • Unquoted keys
  • Comments inside JSON
  • Trailing garbage after the closing brace
If another tool accepts these, it is using a lenient parser. The server consuming your JSON almost certainly uses a strict parser — so Rawtell's verdict is the one to trust. Paste the text in Rawtell, read the line/column message, and fix that byte.
Rawtell says "host unreachable" — is my API down?
Not necessarily. Rawtell runs on Cloudflare Workers. A host that is itself behind Cloudflare cannot be reached from a Cloudflare Worker. This is a platform limitation, not a statement about your API's health. If your API is on Cloudflare, try testing from a non-Cloudflare network or use the Parse text tab with a saved response body instead.
Does Rawtell store the JSON I paste or the URLs I submit?
No. Pasted JSON text is parsed entirely in your browser and never transmitted to any server. Submitted URLs are fetched on the server to retrieve the response, then the URL and response are returned to your browser and discarded — they are not stored, logged, or retained. Rawtell has no user accounts, no databases, no analytics, and no tracking. See the Privacy policy for details.
The URL inspection returned HTML even though the endpoint is supposed to return JSON — what happened?
This is exactly the scenario Rawtell is designed to diagnose. A server that returns content-type: application/json with a 200 OK status and an HTML error page is a common misconfiguration. Rawtell flags the mismatch:
  • It checks whether the body actually looks like JSON (starts with { or [)
  • It reports what the content-type header claims
  • It shows the first few hundred bytes of the body so you can see the HTML yourself
If the server returned a non-2xx status, every derived verdict is marked not established — Rawtell does not guess whose problem it is.
How many redirects does Rawtell follow?
Up to 5 redirect hops. If the chain is longer, Rawtell reports the redirect that exceeded the limit and shows the chain up to that point. The tool fetches only the URL you type — it does not follow links, crawl, or retry failed endpoints beyond the initial redirect chain.
Can I use Rawtell to test internal APIs, localhost, or private IPs?
No. Rawtell blocks URLs pointing to loopback addresses (127.0.0.1, localhost, ::1), private IP ranges (10.x.x.x, 172.16-31.x.x, 192.168.x.x), and URLs containing embedded credentials (https://user:pass@host/). This is a security measure — the server-side fetcher should not access internal networks on your behalf. Use the Parse text tab with a saved copy of the response body instead.

Known limitations

Limitation

Cloudflare-hosted targets cannot be inspected

Because Rawtell runs on Cloudflare Workers, any target that is itself behind Cloudflare reports as unreachable. This is a platform restriction that cannot be worked around from within the Worker runtime. The limitation is noted next to the URL input on the App page.

Limitation

No user accounts, saved history, or dashboards

Rawtell is a stateless, single-use tool. Each inspection is independent. There are no user accounts, no login, no saved-history feature, and no dashboard. Pro and Team plans (with planned saved history and API access) are described on the pricing page but are not yet available — no payment processor is connected yet.

Limitation

No email, alerts, or notifications

Rawtell does not send email, alerts, reminders, or any form of notification. There is no email sender in this fleet. Any page or feature description implying email delivery (such as "email alerts on schedule changes" on the Pro plan) is a description of a planned capability, not an available feature.

Limitation

No payment processor connected

The Pro ($8/month) and Team ($24/month) plan prices are stated as intended pricing. No payment infrastructure is operational. The checkout form on the pricing page is a demonstration only. See needs.md in the repository for what the owner needs to do to connect Stripe.

Limitation

No legal entity registered

The service has no registered company, registered address, or governing jurisdiction. The privacy policy and terms of service are published as statements of operational intent and are not yet legally binding contracts. This will be resolved when the owner registers the business.

Contact us

There is a built-in feedback channel on the App page — scroll to the bottom of the page. Submissions go into a database that someone reviews.

Available channels

In-app feedback form — scroll to the bottom of /app. Select the type (bug report, feature request, general), write your message, and optionally leave a contact handle. This is the only working channel.

Email, ticket system, Discord, X: none of these exist yet. There is no mailbox on any domain associated with this service. If you provide a contact handle in the feedback form, you may hear back — but the operator has no obligation or guaranteed response time, and no automated reply system exists.

For abuse reports or legal concerns: Because no legal entity is registered and no mailbox exists, there is currently no formal channel for legal correspondence. These are honest operational facts about this service, not evasions.