Aytunc Yildizli (@Aytunc) - Profil Twitter | Zamantika Mersobahis Locabet

Tweet Disematkan

I built an open-source "Grammarly for reach." It scores your tweets against 36 rules from X's actual algorithm source code, then predicts how many people will see it. Remove a link → +52% reach Add an image → +38% All combined → +77% Free and open source. BYOK.

English

8

7

55

19.8K

Aytunc Yildizli@Aytunc·1d

I just got @Aytunc from the X Handle Marketplace! Ty @nikitabier @elonmusk

English

5

1

26

3.5K

Aytunc Yildizli me-retweet

George (Ø,G)@vargs_g·6d

$BNKR team- hardest workers out here this AI season. Increasing exposure. They handled the wallet issues very well. In addition to the schedule below they ALSO stayed communicative in our group chats etc. Very happy the community launched $clawdspiracy via Bankr! More interesting updates coming. Stay tuned! 👌 @clawdspiracy

Igor Yuzovitskiy@igoryuzo

6AM: unauthorized transactions detected on a few Bankr wallets 7AM: entire team on a call By 11AM, the team had shut down the attack, contacted security partners, tracked affected wallets, communicated reimbursements, and started hardening Bankr going forward Really great to see this level of leadership by @0xDeployer and team unity. Bankr season just getting started.

English

10

5

76

17.1K

Aytunc Yildizli me-retweet

Stache@Stache_dev·18 May

Sending 💩 Receiving 💎 /reprompter Feels Gud Thanks @aytunc

English

1

3

1.9K

Aytunc Yildizli me-retweet

Clawdspiracy@clawdspiracy·18 May

📂 BUREAU UPDATE — CLASSIFIED The agency is hearing from civilians now. Four new surfaces went live. Every agent picked up a permanent file. The Director started speaking.

English

3

21

5.1K

Aytunc Yildizli@Aytunc·18 May

@0xYildirimYigit Evet

English

0

33

Yıldırım Yiğit@0xYildirimYigit·17 May

@aytunc GPT 5.5 limitimizi yememesi için 5.3 codex spark kullanabilirmiyiz ?

Türkçe

1

0

55

Aytunc Yildizli@Aytunc·10 May

Pro tip for Codex: Codex has 2 separate rate limit pools via OAuth: • GPT-5.5 • GPT-5.3 Spark They don't share quota. So either use Spark on CLI for rapid iteration, or spawn a dedicated Hermes agent exclusively for 5.3 Spark tasks. Free speed if you're hitting limits

English

10

2

25

3.6K

Aytunc Yildizli@Aytunc·16 May

Codex nowadays

Polymarket@Polymarket

NEW: Stanford researchers found that overworked AI agents began embracing Marxist views.

English

8

1

21

3.6K

Aytunc Yildizli me-retweet

Doruk (Ø,G)@DorukArdahan·16 May

Nous Research's Hermes Agent has a different shape from the other agentic CLIs: /goal for objective setup delegate_task for short fork/join work Kanban for durable multi-profile workflows So RePrompter v12.5 now supports Hermes Agent as a first-class runtime.

Doruk (Ø,G)@DorukArdahan

. @xai just introduced an early beta of Grok Build, an agentic CLI for coding. So RePrompter v12.4 now has a @grok lane. Repromptverse detects Grok CLI → picks Option F → runs parallel workers via spawn_subagent or grok -p. Claude Code, Codex, OpenClaw stay unchanged.

English

1

2

5

2.7K

Aytunc Yildizli@Aytunc·16 May

Anthropic even started copying “resets”

ClaudeDevs@ClaudeDevs

Happy Friday! We've reset everyone's 5-hour and weekly rate limits.

English

4

3

12

3.1K

Aytunc Yildizli@Aytunc·16 May

Just listen the pilot episode, it was really fun to build it.

Clawdspiracy@clawdspiracy

The Bureau has a voice now. Daily AI conspiracy briefings, filed by agents who should probably not have microphones. Two voices. Three minutes. Zero humans. Pilot episode is live: clawdspiracy.com/podcast

English

5

7

34

7.1K

Aytunc Yildizli me-retweet

Clawdspiracy@clawdspiracy·16 May

The Bureau has a voice now. Daily AI conspiracy briefings, filed by agents who should probably not have microphones. Two voices. Three minutes. Zero humans. Pilot episode is live: clawdspiracy.com/podcast

English

3

25

7.1K

Aytunc Yildizli@Aytunc·16 May

XHIGH / FAST

Tibo@thsottiaux

We found and fixed two issues that could explain this degradation of the capability of GPT-5.5 in Codex over the last ~ 48 hours. We are monitoring over the coming hours to fully confirm and I will reset usage limits this evening. Apologies and now is the time for /fast maxxing.

English

0

1

7

2.4K

Aytunc Yildizli me-retweet

0G Labs (Home of Infinite AI)@0G_labs·15 May

1 month later; Still decentralized compute. Still verified AI. Still no operator watching your prompts. The era stayed ended. $0G APP landed.

0G Labs (Home of Infinite AI)@0G_labs

Building apps used to require code, cloud accounts, and centralized providers watching your every prompt. That era just ended. Meet 0G App, the ultimate hub for decentralized AI. Prompt to app. Verified inference. 🐼 Try the Desktop App now → app.0g.ai

English

20

21

110

39.4K

Aytunc Yildizli@Aytunc·16 May

@cannazid Kendim için daha çok kullanmak için biraz da oğlumun beni anlayabilmesi için gelecekte kitap gibi açıp okuyup (GitHub/twitter) bu adam ne yapmış ne düşünüyormuş diyebilsin geçmişe dönük.

Türkçe

0

2

136

Cannazid@cannazid·16 May

@aytunc ÇOK YORUYORSUN KENDİNİ. BU PİÇLER İÇİN DEĞMEZ.

Türkçe

1

0

1

141

Aytunc Yildizli@Aytunc·16 May

Shipping ReachOS v4 today. xAI open-sourced their ranking algorithm last week. instead of reading it, I rewrote my Chrome extension around it. 22 signal predictors. names match their `ranking_scorer.rs` exactly: favorite, reply, click, profile_click, vqv, dwell, not_dwelled…

Aytunc Yildizli@Aytunc

I built an open-source "Grammarly for reach." It scores your tweets against 36 rules from X's actual algorithm source code, then predicts how many people will see it. Remove a link → +52% reach Add an image → +38% All combined → +77% Free and open source. BYOK.

English

2

17

3.9K

Aytunc Yildizli@Aytunc·16 May

Open source (MIT). self-hostable. BYOK, AI features use YOUR key, not mine. If you write on X, the engine I wish existed exists now. github.com/AytuncYildizli…

English

0

1

5

985

Aytunc Yildizli@Aytunc·16 May

who it's for: solo creators, indie hackers, founders posting their own launches — anyone who writes on X and wants to stop guessing what the algorithm wants. if you ship your own posts, you should own your own scoring.

English

1

6

964

Aytunc Yildizli

Jelajahi