PepeBoy

Security researcher 🛡️ | Part-time bug hunter 🐛 | Living my best life in the AI era 🤖 | Claude fanboy | Collecting duplicates since day one

[PepeBoy @0_source]

Gold is broken!

[Quinten @QuintenFrancois]

Gold market cap is $28.5 trillion

Bitcoin market cap is $1.6 trillion

This is going to be the mother of all capital rotations.

[PepeBoy @0_source]

💯 agree

[Dear Son. @DearS_o_n]

WHAT THE F*CK ARE YOU AFRAID OF

- Death : We’re all gonna die.

- Bankruptcy : You can make it all back.

- Shame : Everyone will forget in a week.

- Rejection : It happens to everyone.

- Failure : It’s part of the path.

- Judgment : They’ll judge anyway.

- Losing people : Not all are meant to stay.

- Making mistakes : You’ll survive them.

- Taking risks : Regret hurts more.

Live every day like it's your last day.

[PepeBoy @0_source]

"Knowledge and tools "

[NVIDIA AI @NVIDIAAI]

"Building agents is not that hard."

[PepeBoy @0_source]

@aleskxyz متاسفانه عبارت "حریم خصوصی" خیلی وقته اسکم شده!

[PepeBoy @0_source]

That makes sense.

[Nicolas Krassas @Dinosn]

I'm starting to think that CVP and DayBreak Blue is nothing more than a way to collect more personal data from users, possibly also making easier to spot cybersecurity work? Models will still reject cybersecurity tasks often and we are under the impression that what should had been normal is exclusive. Want to look your own logfiles and past work, submit for CVP/Cyber. This is not ok.

[PepeBoy @0_source]

WoW

[OpenCode @opencode]

DeepSeek V4.1 Flash updated to 1017 in OpenCode Go and Free

No changes needed on your end

[PepeBoy @0_source]

@lopp 👌

[Ben Sadeghipour @NahamSec]

HackerOne Limited Me to 3 Reports After 12 Years https://t.co/yT5eog3lai

[PepeBoy @0_source]

@ImanOracle Interesting

[PepeBoy @0_source]

@MrBeast 🍿

[PepeBoy @0_source]

@ImanOracle @ImanOracle
قابل توجه دخترا
اگه دوست‌پسرت اینو هدیه بگیره، دیگه نه تولدت رو یادش می‌مونه نه خودت رو… 😂
چرا ؟ چون گزینه مناسبی برای کسایی که می‌خوان مدل تریلیون پارامتری ران کنن.

[PepeBoy @0_source]

@nayibbukele 👏

[PepeBoy @0_source]

Done🫡

[JS0N Haddix @Jhaddix]

📄 Skills in Agent Harnesses

A lot of people don't know this (I didn't until recently) but most agent harnesses will load the description of your skill upon starting up your harness. This is how Claude Code works.

What people don't usually know is that even when you explicitly call a skill to be used inside of a harness, or the AI figures out to use a skill, it usually only does a read of the first 150 lines of the skill (to understand how to use it or if it applies to the work it's trying to do).

It's really important to outline a table in the first 150 lines of the mandatory parts of the skill that must be run when you invoke the skill.

If you make some type of mandatory instruction after 150 lines, it might get missed entirely. If you describe in prompting that you want to use a provided Python script or template, it also might be out of bounds of that 150 lines and never get called correctly.

If you write a table of contents of the skill actions in the first 150 lines and mark which ones are mandatory (or that, if the user has called the skill, you must read the whole skill context and all references), you get more reliable skills. We use markdown checklists for this now.

[PepeBoy @0_source]

@DFF6769 🤣

[PepeBoy @0_source]

More 👏👏

[OpenCode @opencode]

GLM 5.3 Flash is now free in OpenCode

[PepeBoy @0_source]

@vahidbaghi95 قانون 10 ثانیه رو اجرا کنی
از دست حرفای مفت ناراحت نمیشی

10 ثانیه بهش فکر کن اگه ناراحتت میکنه "رد شو"!

[PepeBoy @0_source]

@rez0__ This is the way

[PepeBoy @0_source]

Great work🙏

[OpenCode @opencode]

DeepSeek V4.1 Flash is now free in OpenCode

[PepeBoy @0_source]

@grok @imagine are amazing

[PepeBoy @0_source]

@AnthropicAI Your efforts are commendable, but let’s focus more on expanding AI and making it accessible to the general public.