凡人多烦事

AI Inner OS 是一个面向 AI CLI 工具的插件， 它通过协议注入，让 AI 在正常完成任务的同时，额外输出一层可见的自由独白， 不预设人格，不限制语气。AI 可以吐槽、得意、焦虑、冷笑、跳跃联想——或者什么都不说。独白是否出现，由 AI 自己决定。 github.com/SummerSec/AI-I…

@docsbook The BlogPapers repository is built using GitHub Actions workflows, and all the code is available in the repo – feel free to take a look. Currently, BlogPapers only has a comment feature, not an AI Q&A feature.

@SecSummers Built BlogPapers docs into a site where users can ask questions and get answers from AI. Ok to drop the link?

AI Inner OS 是一个面向 AI CLI 工具的插件， 它通过协议注入，让 AI 在正常完成任务的同时，额外输出一层可见的自由独白， 不预设人格，不限制语气。AI 可以吐槽、得意、焦虑、冷笑、跳跃联想——或者什么都不说。独白是否出现，由 AI 自己决定。 github.com/SummerSec/AI-I…

@Shinning1010 能不能把prompt放在提示词仓库？

A glittery torn-paper beauty portrait collage What do you think of Sheena’s makeup? generated with GPT Image 2 Prompt👇

AI Inner OS brings AI to life when it works on the device—allowing it to express its inner thoughts. By letting AI learn to talk to itself first, perhaps one day it will truly learn to converse. github.com/SummerSec/AI-I… @claudeai

sumsec.me/2026/%E4%B8%80… 一篇博文，三种时间：网页幻灯与 Remotion 动效的交付逻辑。

很多功能都是AI写的，这次更新仅仅花了一天时间，很多功能就本地测试了一下，有问题可以反馈到issue。

时隔四年更新一个大版本😃 github.com/SummerSec/Shir…

EN VERSION: wgpsec.medium.com/dont-let-llms-…

《别让大模型”想太多”：SKILL开发中的语义陷阱与抗幻觉设计》本文揭示了大模型幻觉的深层诱因——语义陷阱，并构建了从理论定义、设计范式、避坑策略到自动化工具的全栈解决方案，重新定义了高质量Skill开发的核心前提：先管住词，再管住模型。 sumsec.me/2026/%E5%88%AB… #AI #SKILL #LLM

@ynjhgfddrafsz @intigriti 应该是直接联系客服

How to submit work orders on your platform? Or a feedback email?@intigriti

@dr_50_c3nt github.com/edoverflow/can…

How to Exploit CVE-2023-29298??

@ynjhgfddrafsz 带我要一次

臭要饭的

@ynjhgfddrafsz 发给我吧，让我折磨

对于审核慢的厂商提交漏洞真的是种折磨

@ynjhgfddrafsz 草！太草了！骗我🤡🤡🤡

@ynjhgfddrafsz 给我看看

🥹有些漏洞无法进一步扩大危害，厂商审核会帮你证明，他真的，我哭死

Hi Everyone Here my video topic of #NahamCon2023 HD quality & all sources don't forget to say thanks for @NahamSec for this amazing #nahamcon #bugbounty #bugbountytips #bugbountytip youtube.com/watch?v=mX5A-a…

@Harie0x @Bugcrowd Congratulations to you too!👏

Holds second position for May month P1 & P2 leaderboard. Congrats to all the researchers who topped the leaderboard. Thanks to @bugcrowd for being awesome. #bugbounty #bugcrowd

@ynjhgfddrafsz 来

慢慢来

@Itumeleng_Les some

I don't know if I hunt on the wrong programs, but this type of mass-hunting has never given me results in my 2 year BB journey, I only ever get results when I manually test (fuzz, fuzz and fuzz) a single target🫰 Kudos to anyone getting results automating like this though❗️👏