skip to content

Decoder, Comparer & Sequencer

Decoder unwraps layered URL, HTML, Base64 and hex encodings, Comparer diffs two messages by words or bytes, and Sequencer rates token randomness. Interviewers use them to test careful observation.

on this pageshow

explore

questions

6

In Burp Suite's Decoder, how do the Decode as, Encode as, Hash and Smart decode operations differ, and how does it show each layer?

level: juniorimportance: must knowfreq 32%

answer

  1. a chain of panels
  2. four operations on one tab
  3. heuristic first pass
  4. colour marks the encoding type
  5. hash list follows the Java platform

basics

~20 s

Decode as and Encode as apply one chosen format such as URL, HTML, Base64 or GZIP, Hash applies a hash function, and Smart decode peels recognisable layers automatically. Every transformation opens a new colour-coded panel, so each layer stays visible.

solid answer

~40 s

Decoder transforms data you send to it from another tool's message editor or paste in. **Decode as** and **Encode as** apply one function you choose: URL, HTML, Base64, ASCII hex, Hex, Octal, Binary or GZIP. **Hash** applies a hash function, and which functions are offered depends on your Java platform. **Smart decode** looks for encoded data and keeps applying decoding layers until nothing recognisable remains, which makes it a good first step. Each operation opens a new panel with the result, colour-coded by the kind of encoding applied, and any part you did not transform is copied through raw. Because every layer has its own panel, a wrong heuristic guess is easy to spot and back out of.

go deeper

for a junior

Name the four operations and say each result opens in a new panel. You are showing it is a chain you build up, not a one-shot converter.

for a middle

Explain Smart decode as a heuristic that applies layers until nothing recognisable remains and can mistake; the colour-coded per-layer panels are how you catch and correct it.

for a senior

Write the chain into the report: which functions, in which order, and which steps the heuristic chose. A reader who cannot reproduce your decode cannot weigh your finding.

for a principal

Where a standard governs how assessments are written up, require the transformation chain for any value described as encoded or opaque, since that description often carries a severity.

## What Decoder is for Burp Suite's **Decoder** is a utility tab that transforms data with common encoding, decoding and hashing functions. You send it a value from the message editor of another Burp tool, for example a header or a cookie in the HTTP history, or you type or paste data straight into its editor panel. By the documentation's description it works on data you hand it rather than talking to the application under test, so for an assessor holding a written scope, or a reviewer reading a report, it is a reading aid: it shows what a captured value contains and nothing else. ## The four operations | Operation | What it does | Note | |---|---|---| | **Decode as** | Applies one chosen decoding function | URL, HTML, Base64, ASCII hex, Hex, Octal, Binary, GZIP | | **Encode as** | Applies one chosen encoding function | The same list of formats | | **Hash** | Applies a hash function to the data | The functions offered depend on the capability of your Java platform | | **Smart decode** | Looks for encoded data and applies layers of decoding until no further recognisable format remains | A heuristic first step, not a guarantee | You can view the data in either **Text** or **Hex** form; the hex view is the honest one when a decoded layer turns out to be binary rather than text. To transform only part of the data, select that section first. Otherwise the transformation applies to the whole data set. ## Layers, and how Decoder shows them Real values are often wrapped more than once. A parameter may be Base64-encoded so that it is safe to carry, then URL-encoded so that it survives a query string, and the document inside may be compressed on top of that. Decoder is built for exactly this. Every transformation opens a **new editor panel** holding the result, and you can transform that panel again, building a chain from the original value to its final form. Three details matter when you read a chain: - The data is **colour-coded** to indicate which kind of encoding or decoding was applied. - Any part of the data that was **not** transformed is copied into the new panel in its raw form, so nothing is silently dropped. - Every layer lives in its own panel, so you can find the step where a result stopped making sense and restart from the last good panel. ## A worked reading Suppose a reviewer is checking a staging build at `example.internal` and finds a long query-string value in the HTTP history. They send it to Decoder and choose Smart decode. Three panels appear: the first is the URL-decoded text, the second is the Base64-decoded bytes, the third is a GZIP-decoded document that reads as structured text. What the reviewer can now state is factual and limited: the application ships a compressed, encoded document through the client, and anyone who holds the value can unwrap it the same way. Whether that is a finding depends on what the document contains and on what the server later does with it. Decoder supplied the reading, and the judgement still has to come from somewhere else. ## Where the reading stops being trustworthy Smart decode is the convenient part and also the weakest. PortSwigger's page says that it uses heuristic techniques to recognise common encoding formats and can therefore make mistakes. Three habits follow: 1. Read every panel in the chain, not only the last one. 2. Treat a surprising final result as a hypothesis, and confirm the suspect layer with an explicit **Decode as**. 3. Record the chain, meaning which functions in which order and which steps Smart decode chose, so another reviewer can reproduce what you saw. A hash is a different category from an encoding. Hash is not an unwrapping step: there is no layer to undo, and it is there to produce a digest of data you already hold. Because the offered functions depend on the Java platform that runs Burp, two testers on different machines can see different lists, which is a setup fact and not a difference between Burp releases. ## What a defender or reviewer takes from it - **What the value really contains.** A readable decoded form shows what the application put into something any holder can read. - **Which layers were there.** Layered encodings are why a log search for a literal string can miss a value; an engineer lining a report up against a log should know which layers a captured value carried. - **How certain the chain is.** Where Smart decode did the work, say so in the report; where the reviewer chose each step, say that instead. - **What Decoder is not.** It does not judge whether a value is safe, and it makes no request to the target by itself. The only traffic the app owner sees is whatever the Proxy captured earlier. ## Common slips - Reading only the last panel and missing that Smart decode took a wrong turn two layers earlier. - Assuming the Hash list is the same on every machine. - Treating a successful decode as a finding, when the finding is what the server trusts in the decoded content. - Forgetting to select only the relevant part of a message first, so that unrelated text is transformed with it.

  • Why does Decoder show each Smart decode layer in its own panel instead of returning only the final result?
    Smart decode is heuristic and can mistake one encoding for another. Separate, colour-coded panels let you see which layer was applied at each step, spot a wrong guess, and restart from the last good panel instead of trusting an opaque end result.
  • A reviewer expects a particular hash function in Decoder's Hash list and does not find it. What explains that?
    The functions Hash offers depend on the capability of the Java platform Burp is running on, so the list is a property of that installation's Java environment. It is not a statement about the Burp release, and another machine may show a different list.
  • Does using Decoder send anything to the application being assessed?
    Not by the documentation's description: Decoder transforms data you send to it or paste into its editor. The application owner's logs show only whatever the Proxy captured earlier, never the decoding itself, which is why Decoder needs no separate scope consideration beyond handling the captured data with care.

  • Treats Smart decode output as exact without reading the panels it produced
  • Believes Decoder sends the transformed value on to the target
  • Assumes the Hash list is identical on every Burp installation
  • Thinks only one transformation can be applied to a value at a time
  • Calls a successful decode a vulnerability without asking what the server trusts
open as a page

In Burp Suite's Comparer, how do the Words and Bytes analysis options differ, and what can each one miss or bury?

level: middleimportance: should knowfreq 22%

basics

~20 s

Words splits both items on whitespace and reports token-level edits that turn the first into the second. Bytes reports byte-level edits at much higher computing cost, so it is the fallback when Words misses the relevant difference.

open as a page

In Burp Suite's Decoder, what does a cookie that unwraps into readable fields show, and what does one that will not decode show?

level: middleimportance: should knowfreq 26%

basics

~20 s

A cookie that decodes into readable fields is merely encoded: URL, Base64 and hex have no key. A value that will not decode shows only that Decoder did not recognise it, not that it is random or protected.

open as a page

In Burp Suite's Sequencer, how do live capture and manual load differ, and what token sample sizes does the documentation give?

level: middleimportance: should knowfreq 27%

basics

~20 s

Live capture reissues one request and extracts the token from each response; manual load takes tokens you already hold, newline-delimited. In Burp Suite 2026.9, analysis needs 100 tokens, about 5,000 is usually sufficient, and 20,000 is the maximum.

open as a page

A Burp Suite Sequencer analysis of 5,000 session tokens reports strong effective entropy; what does that not prove, and what does it still tell a reviewer?

level: seniorimportance: should knowfreq 18%

basics

~20 s

It is an indicative statistical result, not proof of unpredictability: deterministic output, such as a hash of a sequential number, can pass while anyone who knows the generator's state can extrapolate it forwards and backwards.

open as a page

How do Burp Suite Sequencer's character-level and bit-level analyses differ, and why might the character-level tab be empty?

level: middleimportance: nice to knowfreq 9%

basics

~20 s

Character-level tests (count and transition analysis) examine each position of the raw token; bit-level tests convert tokens to bits and run FIPS, spectral, correlation and compression tests. Character-level tests are disabled at low sample sizes as unreliable.

open as a page