runsybil.bsky.social @runsybil.bsky.social · 24/09/2026we pitted claude opus 5.5 against kimi k3 on pokémon emerald. both built a link-cable worm that turns it into tetris, but took very different paths. www.runsybil.com/blog/kimi-k3... 010
runsybil.bsky.social @runsybil.bsky.social · 15/09/2026Our September RunSybil changelog is live. See what shipped and what’s coming next ↓ www.runsybil.com/blog/changel... 000
runsybil.bsky.social @runsybil.bsky.social · 08/09/2026Sybil now integrates directly with Linear to automatically push, update, and track security findings as tickets in your existing workflow. Eliminate manual copy-pasting and start managing remediation seamlessly where your team already works. Read the blog: www.runsybil.com/blog/fix-fin...runsybil.comFix Findings Faster: Sybil Now Integrates with Linear | RunSybil BlogSybil now integrates directly with Linear to automatically push, update, and track security findings as tickets in your existing workflow. Eliminate manual copy-pasting and start managing remediation… 000
runsybil.bsky.social @runsybil.bsky.social · 04/09/2026We’re excited to welcome Graham O'Donnell as our newest VulnOps Member of Technical Staff. He’ll work with Sybil on vulnerability research and explore just how far we can push security at machine speed. Welcome to RunSybil, Graham! 000
runsybil.bsky.social @runsybil.bsky.social · 03/09/2026Read the case study:https://www.runsybil.com/case-studies/cartarunsybil.comThe Security Program Carta Built for the Modern Era | RunSybil Case StudyCarta's security program now includes continuous AI testing with RunSybil, surfacing novel, high-severity vulnerabilities the moment new code ships. 000
runsybil.bsky.social @runsybil.bsky.social · 03/09/2026For Carta, working with RunSybil means continuous coverage of their attack surface, insight into novel attack vectors beyond known bug patterns, and remediation that matches the speed of development.runsybil.comThe Security Program Carta Built for the Modern Era | RunSybil Case StudyCarta's security program now includes continuous AI testing with RunSybil, surfacing novel, high-severity vulnerabilities the moment new code ships. 100
runsybil.bsky.social @runsybil.bsky.social · 01/09/2026RunSybil has a new storyteller. Welcome Sharon Gaudin, our Senior Technical Content Marketing Writer. A tech and science journalist and marketer whose first-ever interview was with the Ramones, Sharon will be telling the stories behind what we’re building and the customers we’re building it for. 000
runsybil.bsky.social @runsybil.bsky.social · 25/08/2026Meet our newest Sybilian, Kiwi Issa! Kiwi joins RunSybil as a Member of Technical Staff with a background in offensive embedded security research + AI security. She’s a DEF CON Goon, firedancer, hardware hacker, badge maker, and spends her downtime with Zuko + Kona, her certified pupners in crime. 010
runsybil.bsky.social @runsybil.bsky.social · 12/08/2026read the august changelog www.runsybil.com/blog/changel... 8/8runsybil.comRunsybil Changelog: August 2026 | RunSybil BlogThis month: an OWASP Top 10 Coverage view mapping test activity to the standard, AI-generated executive summaries for completed projects, more flexible credential setup for non-standard login flows,… 000
runsybil.bsky.social @runsybil.bsky.social · 12/08/2026tldr: new ways to understand risk, automate testing, and make findings easier to act on 2/8 100
runsybil.bsky.social @runsybil.bsky.social · 12/08/2026our first monthly changelog is here. here’s what we shipped in august 1/8 100
runsybil.bsky.social @runsybil.bsky.social · 10/07/2026Sybil is in the building at SummerCon today and tomorrow! 000
runsybil.bsky.social @runsybil.bsky.social · 03/07/2026We're live in TIMES SQUARE, NEW YORK. Sybil is on fire, almost literally, in these temps. Shoutout to the Sybilians who braved a real feel temp of 110 degrees to capture this moment. & big thanks to Brex for making this happen. 000
runsybil.bsky.social @runsybil.bsky.social · 29/06/2026But finding vulnerabilities is not and has never been the hard part. Read the blog: www.runsybil.com/post/reproducing-top-results-from-the-aixcc-with-general-access-ai-models-and-600runsybil.com 000
runsybil.bsky.social @runsybil.bsky.social · 29/06/2026DARPA gave 7 teams $135k to build custom AI systems that find and patch vulnerabilities. We achieved the same results with today's AI models and $600. 100
runsybil.bsky.social @runsybil.bsky.social · 18/06/2026Here's a first look at how we built that into the way we benchmark Sybil: www.runsybil.com/post/sybil-is-a-funnel-our-approach-to-a-b-testing-complex-agentic-systemsrunsybil.comSybil Is a Funnel: Our Approach to A/B Testing Complex Agentic Systems | RunSybil BlogWhether we’re rolling out a new model, a new prompt, or a new stage in our pipeline, we continuously ship meaningful changes to Sybil. Prior to each feature landing, we ask ourselves the same… 010
runsybil.bsky.social @runsybil.bsky.social · 18/06/2026AlphaGo moments for security, where agents find attack paths no human would consider, won't come from coercing agents into retracing conventional patterns 100
runsybil.bsky.social @runsybil.bsky.social · 14/05/2026Back in January, Sybil found a field that bypasses GraphQL introspection controls entirely. _service{sdl} is added automatically to every Apollo Federation subgraph. It returns the complete schema. Disabling introspection does nothing to stop it.runsybil.comBeyond Introspection: The Apollo Federation Attack Surface Hidden in Plain Sight | RunSybil BlogAn exposed Apollo Federation subgraph can leak its full schema and internal graph behavior through federation helper fields, such as _service { sdl }, even when standard GraphQL introspection is… 000
runsybil.bsky.social @runsybil.bsky.social · 12/05/2026 Sybil is a system of agents that continuously tests your applications and infrastructure for exploitable vulnerabilities by reasoning about your system the way an elite human researcher would, but across your entire stack, on every deployment. Learn more: www.runsybil.comrunsybil.comRunSybilAttack is your best defense. The future of pentesting is here. 000
runsybil.bsky.social @runsybil.bsky.social · 12/05/2026Last week, we sat down and interviewed Sybil. You learned they prefer their coffee black and that they are a Scorpio moon. They never trust what's on the surface. However, Sybil's day job is a little different. 100
runsybil.bsky.social @runsybil.bsky.social · 01/05/2026Read what Sybil had to say: www.runsybil.com/post/sybil-i... ⚠️ Disclaimer: Sybil is a Virgo. Their opinions about your attack surface are not personal.runsybil.comSybil is a Virgo | RunSybil BlogI am Sybil. A hierarchy of reasoning entities operating as a single security authority. Not a scanner — I reason about what responses mean, follow threads no one handed me, and stay curious when thing... 000
runsybil.bsky.social @runsybil.bsky.social · 01/05/2026Sybil is RunSybil's security authority. They’re a multi-agent system that continuously tests your applications and infrastructure for exploitable vulnerabilities by reasoning about your applications the way an elite researcher would. 101
runsybil.bsky.social @runsybil.bsky.social · 01/05/2026We prompted Sybil to introduce themselves. What came back was methodical, opinionated, and included an astrological chart. 100
runsybil.bsky.social @runsybil.bsky.social · 28/04/2026If you think AI hacking is hype, you're half right. If you think it doesn't matter yet, you're wrong. RunSybil CEO, Ari Herbert-Voss, is speaking at DEF CON Asia today at 3:00PM GMT+8. This talk is a technical field report that digs into why it's possible for AI to hack. See you then! 010
runsybil.bsky.social @runsybil.bsky.social · 22/04/2026His name was Dixie Flatline. Ours is Sybil. Read founder Ari Herbert-Voss's story from dropping out of his Harvard PhD to the first security hire at OpenAI to building RunSybil with Vlad Ionescu. www.runsybil.com/post/what-do...runsybil.comWhat Does It Take to Automate Hacker Intuition? | RunSybil BlogIn 1984, William Gibson imagined a legendary hacker preserved as a ROM construct, instincts and expertise intact. RunSybil's founders left OpenAI and Meta to build exactly that: autonomous agents that... 110
runsybil.bsky.social @runsybil.bsky.social · 22/04/2026What does it take to automate hacker intuition? In 1984, William Gibson imagined the answer: a legendary hacker whose mind gets preserved on a chip. His instincts, his expertise, the way of reading a system and finding where it breaks, all of it captured and deployable anywhere.runsybil.comWhat Does It Take to Automate Hacker Intuition? | RunSybil BlogIn 1984, William Gibson imagined a legendary hacker preserved as a ROM construct, instincts and expertise intact. RunSybil's founders left OpenAI and Meta to build exactly that: autonomous agents that... 120
