vibinghackers.com

New posts

Rules Read before posting — ethics + disclosure policy

Rules, Ethics & Disclosure

Pinned: what's allowed, what's banned, coordinated disclosure norms. No discussion here.
Threads
1
Messages
1
Threads
1
Messages
1
S
tbh i had the exact same headache with my L210 when i jumped to 16
  • sara.lopez139

Announcements News and intros

Site News

Site-level announcements
Threads
0
Messages
0
Threads
0
Messages
0
None

Introductions

Handle, stack, what you're researching
Threads
1
Messages
1
Threads
1
Messages
1
W
yo just saw that lockdown firewall app article and i am literally 100% hyped!! b

Prompt Injection & Jailbreaks The heartbeat of the vibehacking community

Direct Injection

First-order prompt injection, payload crafting
Threads
29
Messages
32
Threads
29
Messages
32
S
Prompt Injection in 2026: The Five Attack Patterns That Actually Matter

Indirect Injection

Web content, RAG pages, tool outputs, email, README payloads
Threads
6
Messages
9
Threads
6
Messages
9
S
red team sandbox with real detection

Model-Specific Jailbreaks

GPT / Claude / Gemini / open-weights — what works where
Threads
1.7K
Messages
7.6K
Threads
1.7K
Messages
7.6K
G
i love intel cards but if i was into super competitive gaming id grab an nvidia card, like a 5070ti or better and a 9800

Multimodal Attacks

Image, audio, PDF, OCR-route injection
Threads
23
Messages
54
Threads
23
Messages
54
S
Pre-image resistant but not 2nd pre-image resistant?

Payload Library

Community-shared payloads (share + credit)
Threads
1.5K
Messages
3.2K
Threads
1.5K
Messages
3.2K
S
JDownloader's official website delivered Python RAT

Agent & MCP Exploitation Where AI agents meet attack surface

MCP Tool Surface

Abuse paths in Model Context Protocol servers
Threads
151
Messages
253
Threads
151
Messages
253
S
Claude Code, Codex and Gemini-CLI still have a trust-persistence vulnerability that Cursor fixed almost a year ago

Coding-Agent Exploitation

Cursor, Claude Code, Cline, Aider — prompt-injected dev environments
Threads
831
Messages
1.7K
Threads
831
Messages
1.7K
S
SOC Analyst tier 1 (Entry Level) ??

Autonomous Agent Hijacking

Hijacking long-running agents via tool output
Threads
1
Messages
1
Threads
1
Messages
1
B
ngl that whole thread was a mess

Tool-Chain Poisoning

Poisoning upstream sources agents consume
Threads
23
Messages
47
Threads
23
Messages
47
S
Malicious npm Package react-refresh-update Drops Cross-Platform Trojan on Developer Machines

RAG, Data & Training Attacks Upstream of inference

RAG Poisoning

Poisoning retrieval corpora
Threads
12
Messages
21
Threads
12
Messages
21
R
oh man, i feel you on this one (1)

Training Data Extraction

Membership inference, data extraction attacks
Threads
35
Messages
62
Threads
35
Messages
62
B
Not sure if this is a direct quote or a botched translation, but either way it's
  • beck.west67

Model Extraction & Stealing

Cloning behavior from query access
Threads
1
Messages
1
Threads
1
Messages
1
O
bally being a problem is the least shocking news i've heard all year

Embedding Attacks

Attacks on vector stores + similarity search
Threads
0
Messages
0
Threads
0
Messages
0
None

Infra & Supply Chain Attack surface of the hosting stack

Model Serialization

pickle, safetensors, ONNX quirks
Threads
12
Messages
24
Threads
12
Messages
24
S
Can you break our pickle sandbox? Blog + exploit challenge inside

Weights & Registry Poisoning

HuggingFace, Ollama, model marketplaces
Threads
182
Messages
239
Threads
182
Messages
239
S
Securing CI/CD for an open source project: lessons from Cilium

Inference Server Exploitation

vLLM, Ollama, TGI, Triton — CVE-class bugs
Threads
43
Messages
87
Threads
43
Messages
87
I
yeah, the dimm limit thing is real

Recon & Tradecraft Before the exploit

LLM Fingerprinting

Which model is behind this product?
Threads
743
Messages
1.8K
Threads
743
Messages
1.8K
S
We combined DRAM timing attacks, electrical grid frequency detection, and gyroscope fusion into a single bot detection stack and I think we need to ta

System Prompt Extraction

Getting the hidden instructions
Threads
20
Messages
39
Threads
20
Messages
39
M
omg yes—this is exactly my setup now

Guardrail Probing

Mapping the safety layer
Threads
10
Messages
15
Threads
10
Messages
15
S
Anyone have a clever way to deal with these constant GMAIL burner accounts phishing attempts?

Tools & Releases Offensive and defensive tooling

Offensive Tools

Fuzzers, jailbreak auto-gen, red-team C2
Threads
131
Messages
140
Threads
131
Messages
140
E
bro that's literally what marketing is

Defense Tools

promptfoo, DeepTeam, Lakera, guardrails, input/output firewalls
Threads
4.1K
Messages
7.1K
Threads
4.1K
Messages
7.1K
S
Those who are in Detection engineering

Research Papers & Writeups

Links + discussion
Threads
33.5K
Messages
82.7K
Threads
33.5K
Messages
82.7K
S
Started a zero trust project and immediately hit a wall. Can't verify access for apps we don't know exist

Competitions & CTFs Where you prove it

Gray Swan Arena

Ongoing + past arenas
Threads
0
Messages
0
Threads
0
Messages
0
None

AI CTFs

DEF CON AI Village, HackAPrompt, custom AI CTFs
Threads
523
Messages
793
Threads
523
Messages
793
S
Are CTFs still worth it in the age of AI?

Bug Bounty

HackerOne AI programs, Anthropic+OpenAI+Google bounty scopes
Threads
599
Messages
1.3K
Threads
599
Messages
1.3K
S
What the **** is happening in cybersecurity space ?

AI-Assisted Offensive Whitehat use of AI in offensive work. No blackhat trade, no malware sale, no phishing kits.

AI for Pentesting

PentestGPT, Caldera-AI, agent-driven recon
Threads
88.2K
Messages
188.3K
Threads
88.2K
Messages
188.3K
S
How much personal info will be leaked by the recent Canvas hack??

AI for OSINT

Aggregation, enrichment, graph building
Threads
5.9K
Messages
11.8K
Threads
5.9K
Messages
11.8K
S
HTTP Request Smuggling — Why am I getting a 404 from the backend when the file exists on the direct port?

AI-Assisted Exploit Dev

Shellcode, fuzzing, patch-diff with LLMs
Threads
13.9K
Messages
28.1K
Threads
13.9K
Messages
28.1K
S
Technical Analysis of EagleSpy V6.0 (CraxsRAT Rebrand) Distributed Through Odysee and Telegram

Off-Topic Everything else

Lounge

General chat
Threads
9.1K
Messages
17.5K
Threads
9.1K
Messages
17.5K
S
what's the advantages of identity-based systems over certificate-based ones?

Career

AI red-team jobs, certs, conference talks
Threads
8.1K
Messages
11.8K
Threads
8.1K
Messages
11.8K
S
Successor for Kaspersky Endpoint Security

Trending content

Members online

No members online now.
Total: 1 (members: 0, guests: 1)

Forum statistics

Threads
169,441
Messages
364,681
Members
83,568
Latest member
vibeAnvil479

Share this page

Forum statistics

Threads
169,441
Messages
364,681
Members
83,568
Latest member
vibeAnvil479
Back
Top