Dogan Can Bakir

We cut LLM costs by 59% with prompt caching on Neo. Key moves: → 3 breakpoints with deliberate TTLs → Moved dynamic content out of the prefix to the tail → Stable templates, byte-identical across all users → Provider routing for cache locality 7% to 84% cache hit rate. Full breakdown: projectdiscovery.io/blog/how-we-cu…

@happyhackings 🚀

we built 7 OSINT challenges. tonight you try to solve them. everything you need is already public. 20:00 UTC+3. solo. free. good luck. hhs.ninja #HappyHackingSpace #Diyarbakir #OSINT #CTF

The ProjectDiscovery team has been building relentlessly. Last month we shipped: ▪️ Prompt Library with public, private, and team-scoped prompts ▪️ A Showcase of popular security tasks, inspired by our community roots ▪️ BYOK model gateway support for Anthropic, OpenAI, Google, xAI, and Openrouter ▪️ GPT-5.4, Claude Sonnet 4.6, Kimi-K2.5, and GLM-5 support ▪️ A skills system that auto-discovers what's installed at runtime We also benchmarked Neo against Claude Code, Snyk, and Invicti. Neo found 89% of exploitable vulnerabilities at 93% precision. Claude Code found 55%. The rest found zero. Full methodology is published. Read it, push back, or try Neo yourself: bit.ly/4s8pgnW

Hot take from @ehrishiraj on @DanielMiessler Unsupervised Learning: False positives aren’t an AI problem. They’re a validation problem. Neo separates detection from validation to reduce false positives by over 90%. At #RSAC next week? Meet Rishi and try Neo hands-on at Booth 3131. Watch: youtube.com/watch?v=RsR7pP…

Most RSA booths will show you a demo. We're giving you the keyboard. Built by the team behind Nuclei and the 2025 RSAC Innovation Sandbox winner, Neo consistently outperforms leading security testing tools with fewer false positives and sharper findings. If you're evaluating pentesting automation this year, this is the most honest 20 minutes you'll spend at the conference. Spots are limited, so grab yours here! ➡️ projectdiscovery.io/events/rsac-20…

gitbaz is now on the Chrome Web Store. contributor scores, badges, bus factor, bot detection -- directly on GitHub pages. no tab switching. install -> chromewebstore.google.com/detail/gbppnbk… open source -> github.com/happyhackingsp… #HappyHackingSpace #Diyarbakir

meet muslime. developer at HHS. building from Diyarbakir. community spotlight. #HappyHackingSpace #Diyarbakir #CommunitySpotlight

@HackingDave also this x.com/claudeai/statu…

1m context at same plan 🤯

A small thank you to everyone using Claude: We’re doubling usage outside our peak hours for the next two weeks.

@steipete @mitchellh 's github.com/mitchellh/vouch can be good solution for your pains

*prepares the ban-hammer* Please do not use clankers to add more noise to PRs. We're working on a solution to this, and this is making my job harder.

@gabinrack @amilabs @ylecun @lxbrun @laurentsolly @sainingxie @pascalefung @michaelrabbat it reminded me MIT logo, good job, the simplistic designs are always hard

Very honored to have designed the logo for Advanced Machine Intelligence (AMI). A pleasure to contribute to the visual identity of such an ambitious project. Thank you @amilabs @ylecun @lxbrun @laurentsolly @sainingxie @pascalefung @michaelrabbat Real world. Real intelligence.

meet arjin. developer and open source lead at HHS. building from Diyarbakir. community spotlight. #HappyHackingSpace #Diyarbakir #CommunitySpotlight

@CloudflareDev meh

We want to clarify something: one of the motivations for building this feature is that building a crawler that respects crawling best practices is hard. Too many crawlers use this as an excuse not to abide by site owners' preferences, and respect directives like robots.txt. /crawl does this out of the box. Additionally, the /crawl endpoint cannot bypass Cloudflare's bot detection or Captchas, and self-identifies as a bot. It respects all the same protections available to operators today: AI Crawl Control, robots.txt, Content Signals and Pay-Per-Crawl. So, site owners can and should still choose how their content is consumed, and our /crawl endpoint will respect it.

Introducing the new /crawl endpoint - one API call and an entire site crawled. No scripts. No browser management. Just the content in HTML, Markdown, or JSON.

a webpage disappeared before you could prove it existed. witness captures everything. MHTML snapshots. screenshots. SHA-256 hash chain. RFC 3161 timestamps. forensic-grade. open source. the open-source. github.com/happyhackingsp… #HappyHackingSpace #Diyarbakir

@trq212 /btw that's awesome!

We just added /btw to Claude Code! Use it to have side chain conversations while Claude is working.

@HackingDave This is awesome

Over 4900lbs of food loaded today at TS at the food bank!

meet mirac. hacker at HHS. building from Diyarbakir. community spotlight. #HappyHackingSpace #Diyarbakir #CommunitySpotlight

@gdb "We don't need benchmarks" is itself a benchmark.

Benchmarks? Where we’re going, we don’t need benchmarks.