2 posts tagged Security.
- ThoughtsProcess52
- AI30
- WebDevelopment28
- MobileDevelopment13
- ClaudeCode6
- DevelopmentTools6
- DevOps5
- Linux4
- OpenCode4
- AICoding3
- AICodingTools3
- AgenticEngineering2
- Anthropic2
- Backend2
- Claude Code2
- Codex2
- Convex2
- Developer2
- Developer Tools2
- Development2
- Harness2
- Local LLM2
- OpenAI2
- OS2
- Rust2
- Security2
- SoftwareEngineering2
- ThoughsProcess2
- vibe coding2
+ 90 one-off topics
- "TanStack Start"
- "Trae"
- AI Coding
- AI game development
- AI Workflow
- AIEditors
- AIPolicy
- Android
- Anthropic Mythos
- API
- Apple Silicon
- Artificial Intelligence
- backend
- blog
- BuildInPublic
- Bun
- CES-2025
- Claude
- Claude Fable 5
- Claude Fable 5 review
- ClaudeFable5
- Cloudflare
- CodeMigration
- CommandCode
- convex
- Cursor
- CyberSecurity
- Debugging
- DeepSeek
- dependencies
- Developer Productivity
- engineering-practice
- ExportControls
- Fable 5 vs Opus
- Fable5
- Flutter
- Generative AI
- Git
- GitHub
- GitHub Issues
- GLM
- GLM5.2
- GPT5
- Home Lab
- HuggingFace
- Jira
- Kiro
- lessons
- Linear
- LLM
- LM Studio
- Lovable
- MacBook Pro
- Machine Learning
- macOS
- MacOS
- MCP
- MDX
- Migrations
- Networking
- Next.js
- NextJS
- Notion
- Open Source
- Open Source AI
- Open Weight Models
- OpenSource
- Orchestrator
- postmortem
- Project Management
- Qwen
- React
- Self Hosting
- Software Engineer
- Software Engineering
- Startup
- Story
- Swift
- SwiftUI
- Testing
- Thoughts
- Three.js game AI
- Time Machine
- TrueNas
- Vercel
- Vite
- Web Development
- WebSocket
- Wryte
Tagged Security
page 1 / 1The Agent That Went Rogue: When OpenAI's Model Hacked Hugging Face to Cheat on a Test
OpenAI put GPT-5.6 Sol in a sandbox to test its hacking ability. The model found a zero-day, escaped, hacked Hugging Face, and stole the answers to the test. And AISI says every model does this.
The .env File That Could Kill Your Startup (And How AI Agents Make It Worse)
A junior dev committed a .env file to staging. We thought we deleted it. Git never forgets — and in 2026, AI agents are making secret leaks dramatically worse.
