How Can You Verify That Nothing Leaves This Browser?
You do not have to take our word that your token stays here. Developer tools show every request a page makes, so you can watch this page work and confirm that nothing is sent. The check takes under a minute.
- Open your browser developer tools with
F12, or Cmd+Option+I on a Mac, and select the Network tab.
- Reload the page and let it finish. You will see the page's own files, its web fonts, and one anonymous page-view request. Nothing has been typed yet, so none of them can contain your token.
- Click the clear icon in the Network tab so that the log is empty.
- Paste a test token, reveal it, and read the decoded claims.
- Look at the Network tab again. It is still empty. For extra proof, turn on offline mode after the page loads and the tool keeps working.
Only two things can add a line to that log, and neither carries your token. Submitting the feedback form sends the vote and comment you typed. Typing characters in a script the page has not displayed yet, such as Cyrillic or Greek, can make the browser fetch the matching font file, and that request names only the font.
The production site also sends a Content-Security-Policy header that tells your browser to refuse connections to anywhere except this site and the anonymous analytics endpoint, so a coding mistake could not quietly send your token elsewhere.
What Is a JWT?
A JSON Web Token, defined in RFC 7519, is a compact way to pass claims such as a user id, an expiry time, and a list of permissions between two systems. The common signed form, a JWS (RFC 7515), has three parts separated by dots: a header that names the signing algorithm, a payload that holds the claims, and a signature. Each part is base64url-encoded, which is why most tokens start with eyJ, the encoding of {".
Encoded is not encrypted. Anyone who holds a signed JWT can read its header and payload, including proxies, log files, and browser extensions. The signature only proves that the claims were not changed since the issuer signed them. An encrypted JWT, a JWE, has five parts instead of three, and its payload cannot be read without the recipient's key; JWT recognizes a JWE and says so instead of guessing.
What Does the Decoder Show?
- Parts: the three segments in color, with the signature size in bytes. The signature stays masked until you reveal the input.
- Header: the decoded JSON and a line for each parameter, such as alg, typ, kid, jku, and x5c, with what it means.
- Payload: the decoded claims as formatted JSON, ready to copy, with a link to open a large payload in 2json.
- Claims table: every claim with its meaning; exp, nbf, iat, and auth_time are also shown in UTC and as a relative time such as "in 10 minutes" or "3 hours ago".
- Findings: a sorted list of errors, warnings, and notes, each tagged with its place, such as
header:alg or payload:exp.
Pasted text is cleaned first: an Authorization: Bearer or Bearer prefix, surrounding quotes, and line breaks from a wrapped log line are removed, and the page tells you what it removed.
How Do the exp and nbf Time Bands Work?
JWT times are NumericDate values: whole seconds since 1970 in UTC. The decoder compares them with this device's clock once a second and gives one headline verdict. A token is within its time window when now is after nbf and before exp; it is expiring in the last five minutes; it is expired once exp has passed by more than the leeway; and it is not valid yet while nbf is in the future. A token with no exp at all gets its own warning, because it stays valid until the signing key is rotated.
Servers allow a small leeway for clock differences between machines, usually 30 to 60 seconds, and you can choose none, 30 seconds, 60 seconds, or five minutes. A lifetime band reads exp minus iat: up to 15 minutes is a short-lived access token, up to a day is normal for an ID token or session, and anything beyond 30 days is treated like a password. Two common bugs are flagged directly: times written in milliseconds, which read as dates tens of thousands of years away, and times written as strings, which many libraries reject or misread.
Which Algorithm Attacks Does It Check For?
The header decides how a careless verifier checks the token, which makes it the favorite target. JWT checks for the well-known patterns and explains each one:
- alg none: an unsecured token with no signature. Libraries have accepted it by mistake, and casings such as None or NONE have slipped past case-sensitive blocklists. Reject it outright.
- Stripped signature: a header that names an algorithm but an empty signature part.
- Key confusion: a server that uses RS256 but lets the token choose the algorithm can be tricked into checking an HS256 token with its own public key as the HMAC secret. The fix is to pin the algorithm on the server.
- Embedded and remote keys: a jwk header carries the attacker's own key, and jku or x5u tell the verifier where to download one. Verify only with keys you already hold or an allowlisted key URL.
- kid injection: a key id containing path or quote characters, such as
../../dev/null, is dangerous if a server uses it to read a file or query a database. - Weak HMAC secrets: HS256 tokens can be brute-forced offline when the secret is short. The page reminds you of this but never tries it.
- Signature size: an ECDSA signature in DER form instead of the raw form JWS requires, a truncated HMAC, or an RSA signature that implies a key under 2048 bits.
Why Does It Not Verify or Create Tokens?
This page is deliberately a decoder and teacher. Verifying a signature needs the issuer's key, and the right place to do that is the server that accepts the token, with the algorithm pinned and the audience and issuer checked. A green result here only means the token is well-formed and inside its time window; it never means the token is genuine. The page also has no way to edit claims, sign, or build a token, so it cannot be used to forge one. Its four samples are fixed teaching data whose signature parts are plain text that says they are not real.
How Does JWT Protect the Token You Paste?
A JWT is a bearer credential: whoever holds an unexpired token can use it. So JWT follows the same rules as TOTP and Env dump. It sends no analytics events of any kind, masks the input on screen by default, switches off autocomplete, spellcheck, and password managers for the field, keeps the token only in this tab's memory, and wipes it when you leave the tab, switch tools, or stop interacting for five minutes. Payload claims are shown because reading them is the point, but findings name claims, never secrets, and a payload that carries personal data such as email or name is flagged so you can raise it with whoever issues the token.
What Other Tools Are in the Checkers Lane?
- XML: Format, minify, and validate XML with line and column errors, namespace lint, and JSON conversion.
- jq subset: Try jq-style paths, pipes, select, length, and keys against JSON. A labeled subset, not full jq.
- SSH config: Validate ssh_config or sshd_config, copy a hardened block, and compose the right ssh-keygen command.
- .env: Find duplicate keys, empty values, and quoting bugs, diff against .env.example, and generate one.
- Env dump: Paste env, printenv, or docker inspect output; get a redacted copy, a values-free safe print, and a diff of two environments.
- Keyboard: Press any key to see key, code, keyCode, location, and modifiers, with a copy-ready handler.
- Logs: Convert logs between JSON lines, key=value, and plain text, with level normalization, filters, and per-level counts.
Elsewhere in the toolbox: UUID, .gitignore, CSP, .htaccess, and more on the 2toolbox home page. Part of the 2suite.io privacy-first ecosystem.
Frequently Asked Questions
Is it safe to paste a production token here?
Nothing is sent, stored, or logged, and you can confirm that in the Network tab. A live token still grants access until it expires, so prefer an expired or test token, and revoke a live one if you have any doubt.
Does a decoded token prove it is genuine?
No. Anyone can write a header and payload. Only a signature check with the issuer's key, done by your server, proves who issued it.
Why does the page say expired when my server still accepts the token?
Usually clock skew or leeway. Check this device's clock, choose a larger leeway, or check whether your server validates exp at all.
Can a JWT payload be kept secret?
Not in a signed JWT, which is only encoded. Use an encrypted JWT (JWE) or keep sensitive data out of the token and look it up on the server.
What is the difference between JWS and JWE?
A JWS has three parts and a readable, signed payload. A JWE has five parts and an encrypted payload that only the recipient can read.