1604 matches found
openclaw-technical-analysis
OpenClaw Technical Analysis This repository contains my deep-dive technical reviews and code audits of the OpenClaw AI platform specifically focusing on the February 15 and February 25, 2026 builds. These reviews are grounded in raw architecture analysis, GitHub commit history, and the intersecti...
ActGuard
ActGuard ActGuardã¯ãããŒã«ã䜿çšããLLMãšãŒãžã§ã³ãã«ããã鿥çããã³ããã€ã³ãžã§ã¯ã·ã§ã³ã«å¯Ÿããå®è¡åã¢ã¯ã·ã§ã³ç£æ»é²åŸ¡ã§ãããã®ãªããžããªã«ã¯ãæçµçãªActGuardå®è£ ãšããããè©äŸ¡ããããã«å¿ èŠãªAgentDojoããŒã¹ã®ã©ã³ã¿ã€ã ãå«ãŸããŠããŸãã ãã®ãªãªãŒã¹ã«ã¯ãæçµçãªå®å šãªActGuardæ§æã®ã¿ãå«ãŸããŠããŸããå éšã³ã³ããŒãã³ãã¯ãAgentDojoãã³ãããŒã¯ããã³ãã¬ãŒã¹åœ¢åŒãšã®äºææ§ã®ããã«ãåŒãç¶ãèå¥åreflectiveauditã䜿çšããŠããŸãã èŠä»¶ Python 3.10以é OpenRouter APIããŒ...
PIForge
PIForge RLããŒã¹ã®ããã³ããã€ã³ãžã§ã¯ã·ã§ã³ã¬ããããŒãã³ã°ã®ããã®ãªãŒãã³ãã¬ãŒã ã¯ãŒã¯ ð» ã³ãŒã · ð€ ã¢ã㫠· ð è«æ PIForge㯠PISmith ãš Climbing the Hill ã®å ±æã³ãŒãããŒã¹ã§ããPISmithã¯ããã³ããã€ã³ãžã§ã¯ã·ã§ã³ã¬ããããŒãã³ã°ã«ãããå ±é ¬ã®çæ§åé¡ã«å¯ŸåŠããRLæ»æè ãçšã«æåããæ»æãæ¢çŽ¢ãåŠç¿ã§ããããã«ããŸããPISmithãåºç€ãšããŠãClimbing the Hillã¯ã«ãªãã¥ã©ã åŠç¿ãçšããŠãããå ç¢ãªããã³ãã£ã¢ã¿ãŒã²ããã«å¯Ÿããã¬ããããŒãã³ã°æã®ã³ãŒã«ãã¹ã¿ãŒãåé¡ã«å¯ŸåŠããŸãã âš ãã¥ãŒã¹...
From A2A Attacks to Envelope-Layer Defense: Red-Teaming Evaluation of LLM Agents and a Three-Layer Isomorphic Attack-Defense Model
Agent interaction protocols such as ACP and A2A have moved LLM-based agents toward multi-agent collaboration, introducing new security threats. A task sent by a remote peer over A2A is treated as a legitimate request, providing a natural channel for indirect prompt injection. Existing agent...
CVE-2026-102938
virtualenv is a tool for creating isolated virtual python environments. Prior to 21.7.11, PyEnvCfg.write writes prompt values verbatim to the line-oriented pyvenv.cfg format while PyEnvCfg.readvalues parses the file with str.splitlines and accepts the last value for duplicate keys. An attacker wh...
CVE-2026-102697
Ollama versions 0.14.0 before 0.31.2 contain an incorrect authorization vulnerability in the experimental agent mode Bash tool approval mechanism that fails to properly parse shell syntax. Attackers who can influence model output through prompt injection can execute additional shell commands by...
CVE-2026-102697 Ollama 0.14.0 before 0.31.2 Experimental Agent Bash Approval Bypass via Prefix-Based Authorization
Ollama versions 0.14.0 before 0.31.2 contain an incorrect authorization vulnerability in the experimental agent mode Bash tool approval mechanism that fails to properly parse shell syntax. Attackers who can influence model output through prompt injection can execute additional shell commands by...
CVE-2026-102697
Ollama versions 0.14.0 before 0.31.2 are vulnerable to an incorrect authorization flaw within the experimental agent mode Bash tool approval mechanism . The root cause is a failure to properly parse shell syntax, which allows attackers to bypass session approval requirements. By using prompt inje...
CVE-2026-102697 Ollama 0.14.0 before 0.31.2 Experimental Agent Bash Approval Bypass via Prefix-Based Authorization
Ollama versions 0.14.0 before 0.31.2 contain an incorrect authorization vulnerability in the experimental agent mode Bash tool approval mechanism that fails to properly parse shell syntax. Attackers who can influence model output through prompt injection can execute additional shell commands by...
CVE-2026-102697 Ollama 0.14.0 before 0.31.2 Experimental Agent Bash Approval Bypass via Prefix-Based Authorization
Ollama versions 0.14.0 before 0.31.2 contain an incorrect authorization vulnerability in the experimental agent mode Bash tool approval mechanism that fails to properly parse shell syntax. Attackers who can influence model output through prompt injection can execute additional shell commands by...
EUVD-2026-89089
Ollama versions 0.14.0 before 0.31.2 contain an incorrect authorization vulnerability in the experimental agent mode Bash tool approval mechanism that fails to properly parse shell syntax. Attackers who can influence model output through prompt injection can execute additional shell commands by...
aisecplus-week01-danielossai
ð AISec Plus â 第1é±: è åšããããã³ã°ãã ç ç©¶è : Daniel Ossai Repo: aisecplus-week01-danielossai æåº: 第1é± Â· AISec Plus ããŒããã£ã³ã ð ã€ã³ã·ãã³ãåãšæ¥ä» EchoLeak CVE-2025-32711 çºèŠ: 2025幎1æ · é瀺: 2025幎6æ · ä¿®æ£: 2025幎6æ ð äœãèµ·ãã£ãã Aim Securityã®ç ç©¶è ãã¯ãMicrosoft 365...
CVE-2026-77177
Open GenAI Stack aka ogx-ai 2026-06-11, as used in the Meta AI backend for WhatsApp and other products, allows code execution because prompt injection with Jinja2 template syntax can be used to achieve server-side expression evaluation without sanitization...
langgrinch-cve-2025-68664-analysis
LangGrinch è匱æ§åæ: CVE-2025-68664 æŠèŠ ãã®ãªããžããªã«ã¯ãLangChain Core ã«åœ±é¿ããã·ãªã¢ã©ã€ãŒãŒã·ã§ã³ã€ã³ãžã§ã¯ã·ã§ã³è匱æ§ã§ãã CVE-2025-68664å¥å LangGrinchãåæããæè¡ã¬ããŒããå«ãŸããŠããŸãã æ¬ã¬ããŒãã§ã¯ãè匱æ§ã®èæ¯ãæ ¹æ¬åå ãæªçšãããŒãè åšã¢ãã«ãç·©åçãããã³ LLM ããŒã¹ã®ã¢ããªã±ãŒã·ã§ã³ã«å¯ŸããçŸå®äžçã®ã»ãã¥ãªãã£åœ±é¿ã«ã€ããŠèª¬æããŸãã 察象ããã㯠LangChain ã®ã·ãªã¢ã©ã€ãŒãŒã·ã§ã³ãšãã·ãªã¢ã©ã€ãŒãŒã·ã§ã³ å®å šã§ãªããã·ãªã¢ã©ã€ãŒãŒã·ã§ã³ / CWE-502 LLM...
IPI-exposure-signal
IPIé²åºã·ã°ãã« ããã¯ãç§ãã¡ã®è«æãYour Agentic LLMs Secretly Encode Latent Signals of Indirect Prompt-Injection Exposureãã®ã³ãŒããªããžããªã§ãã ArXivçãšè«æãªã³ã¯: https://arxiv.org/abs/2608.02657 ãã®ãªããžããªã¯ãæœåšçãªIPIé²åºã·ã°ãã«ã®ãããŒãã³ã°ãã€ãã©ã€ã³ïŒAgentDojoãã¬ãŒã¹åéãIPIãªã¹ã¯ã©ããªã³ã°ãé ãç¶æ ã®ãã£ãŒãã£ãŒåãIPIé²åºãããŒãã®ãã¬ãŒãã³ã°/è©äŸ¡ïŒãå®è£ ããŠããŸãã åŒçš...
CVE-2026-77177
Open GenAI Stack aka ogx-ai 2026-06-11, as used in the Meta AI backend for WhatsApp and other products, allows code execution because prompt injection with Jinja2 template syntax can be used to achieve server-side expression evaluation without sanitization...
EUVD-2026-89070
Open GenAI Stack aka ogx-ai 2026-06-11, as used in the Meta AI backend for WhatsApp and other products, allows code execution because prompt injection with Jinja2 template syntax can be used to achieve server-side expression evaluation without sanitization...
CVE-2026-77177
Open GenAI Stack aka ogx-ai 2026-06-11, as used in the Meta AI backend for WhatsApp and other products, allows code execution because prompt injection with Jinja2 template syntax can be used to achieve server-side expression evaluation without sanitization...
CVE-2026-77177
The Open GenAI Stack (also known as ogx-ai ) version 2026-06-11 , utilized in the Meta AI backend for WhatsApp and other products, is vulnerable to code execution . The root cause is a lack of sanitization allowing prompt injection using Jinja2 template syntax , which enables server-side expressi...
skulto
Skulto AIãšãŒãžã§ã³ãã¹ãã«ãã€ã³ã¹ããŒã«ããåã«ãããã³ããã€ã³ãžã§ã¯ã·ã§ã³ãã¹ãã£ã³ããŸã ââââââââââââââââââââââââââââââââââââââââââââââââââââââââââ â âââââââââââ ââââââ ââââââ âââââââââ âââââââ â â âââââââââââ âââââââ ââââââ ââââââââââââââââââ â â âââââââââââââââ âââ ââââââ âââ âââ âââ â â âââââââââââââââ âââ ââââââ âââ âââ âââ â â...