kong

Kong: The Agentic Reverse Engineer LLM orchestration for reverse engineering binaries What is Kong? Most tasks follow a linear relationship: the more difficult a task, the longer it usually takes. Reverse engineering and binary analysis is a task in which the actual difficulty is somewhat trivial...

www-project-top-10-for-large-language-model-applications

www-project-top-10-for-large-language-model-applications OWASP Foundation Web Repository OWASP Top 10 for Large Language Model Applications Welcome to the official repository for the OWASP Top 10 for Large Language Model Applications! About This Repository This repository contains the OWASP Top 1...

inspect_petri

Inspect Petri Welcome to Inspect Petri, an auditing agent that enables automated monitoring and interaction with language models to detect potential alignment issues, reward hacking, and other concerning behaviors. Petri helps you rapidly test concrete alignment hypotheses end‑to‑end. It: Generat...

EntropicaPublic

March updates I need to move some work over from the private repos. The exploration into using massivly multheaded models with mixer and readout functions separated continues. Currently training up a Chinese TinyStories on a 3B-equivalent model that trains, easily, on an 8GB gpu. We're looking at...

markdown-exfil-tester

valtik-markdown-exfil-tester Black-box tester for indirect prompt injection leading to markdown / HTML exfiltration in LLM-backed chatbots. This is the vulnerability class that hit Microsoft 365 Copilot CVE-2025-32711, aka EchoLeak, ChatGPT OpenAI Feb 2026 patch, and Salesforce ForcedLeak. The...

birdy-edwards

BIRDY-EDWARDS 2.0 Infiltrate & Expose Automated AI powered Facebook intelligence tool for target profiling, network analysis and threat reporting. Runs entirely on-device via Ollama. AI-powered Facebook SOCMINT platform — 100% local, zero cloud dependency. For lite version check this repo click...

GenAI-LLM-Top10

OWASP Top 10 for Large Language Model Applications !IMPORTANT Current release: 2026 — published August 4, 2026. Get the 2026 publication · Read the canonical source · Report a correction This is the official source repository for the OWASP Top 10 for Large Language Model Applications , maintained...

oss-fuzz-gen

A Framework for Fuzz Target Generation and Evaluation This framework generates fuzz targets for real-world C/C++/Java/Python projects with various Large Language Models LLM and benchmarks them via the OSS-Fuzz platform. More details available in AI-Powered Fuzzing: Breaking the Bug Hunting Barrie...

deleting-the-trace

Control-token injection attacks on tool-using agents Code and logged measurements for a study of two input-level attacks on tool-using language model agents. The first attack appends a short string of a model's own channel-control tokens to untrusted input; the tokenizer reads it as an...

batch_jailbreak

Official Repo for Safety in Batches? Understanding and Mitigating Safety Failures in Batch Prompting Kihyun Kim, Hee-Seon Kim, Wonjun Lee, Changick Kim Korea Advanced Institute of Science and Technology KAIST News 2026.09 Our paper has been accepted to AACL-IJCNLP 2026 Main! 🎉 2026.08 Paper is...

aegis-audio-defense

AEGIS: Audio Endogenous Guarding via Internal Signals Against Large Audio-Language Model Jailbreaks Submitted to ICASSP 2027. AEGIS keeps the target audio-language model frozen and adds a mid-layer risk gate that conditionally drives late-layer safety adapters, with closed-loop scaling at...

Does Showing Agents the Privilege of Their Tools Reduce Over-Privileged Tool Selection?

When several tools can complete a task, an agent that picks the most powerful one holds more authority than the task needs. Recent work demonstrates that language-model agents do this often, and more so after a tool fails. Approaches to restrict focus so far on either enforcing least privilege...

EleutherAI Language Model Evaluation Harness server-side request forgery vulnerability

The EleutherAI Language Model Evaluation Harness contains a server-side request forgery vulnerability. This vulnerability arises because the lm-evaluation-harness component fails to validate model URLs specified in JobSpecs. Authenticated users can submit malicious model URLs, resulting in the...

Context-Aware Prompt Injection Detection Via Small Language Models in RAG Pipelines: Nokast-SecureRAG

Local Retrieval-Augmented Generation RAG systems are vulnerable to indirect prompt injection and retrieval poisoning, where malicious instructions in retrieved content are mistaken for trusted input. We present nokast-secureRAG, a roughly 0.5B-parameter small language model that acts as a semanti...

Make the Adversary Spend More Energy: Generative Deception against Language-Model Attackers in Critical Infrastructure

Language-model attackers consume computational resources as they explore and interact with a target, creating an opportunity for cyber deception to impose costs. Whether such engagements favor the defender requires measuring both sides. We evaluate an autonomous language-model attacker against a...

DarKnight: Retrieval-Guided Agentic Framework for Breaking Dark Web CAPTCHA Gates

Cyber Threat Intelligence CTI investigators who monitor dark web marketplaces for leaked credentials and breach listings must pass the diverse CAPTCHAs that operators deploy to block automated collection. Existing solvers either need unscalable per-site fine-tuning or rely on commercial...

EUVD-2026-96609

JeecgBoot through 3.9.5 contains a missing authorization vulnerability in the promptExperiment handler of AiragPromptsController that allows any authenticated user to run AI prompt experiments. Low-privileged attackers can supply other users' prompt template and dataset ids to trigger large...

FinRT: Distilling Adaptive Red-Teaming Strategies into Reusable Adversarial Generators in Consumer Finance

In regulated industries like consumer finance, seemingly harmless user queries can exploit large language model vulnerabilities, triggering safety failures and pushing responses dangerously close to policy limits. Existing automated red-teaming methods trade off attack effectiveness against...

SEAD: A State-Based Perspective on Attack and Defense in Tool-Using Agents

Language-model agents increasingly use tools to act on external systems. Earlier actions can alter files, permissions, database records, or other state, making a later routine-looking action harmful. Yet the visible interaction may not reveal the underlying state needed to assess that action. We...

API Secrets Should Never Become Tokens in the LLM's Vocabulary

Tool-using large language model LLM agents turn credential hygiene from a storage problem into an execution-security problem. A key pasted into a prompt, or embedded in a system prompt or tool configuration, crosses from an authentication boundary into a data pipeline, where it may persist in...