NY-squared AI
NY-squared AI @NYsquaredAI ·
Arcjetがインライン プロンプトインジェクション防御を 正式リリース。 500以上の本番アプリで稼働中。 本番AIシステムの防御が 標準装備になる時代。 #PromptInjection #DevSec
5
NY-squared AI
NY-squared AI @NYsquaredAI ·
BREAKING: 3 critical vulnerabilities in LangChain/LangGraph disclosed. CVSS 9.3 deserialization flaw leaks API keys and env secrets. 52M+ weekly downloads affected. AI framework supply chain security is no longer optional. Patch now. #AISecurity #LangChain #DevSec
16
CyberSecurity88
CyberSecurity88 @CSec88 ·
Hot take: AI coding tools are a gift to attackers. 10× more vulnerabilities shipped. 45% vuln rate. 97.5% of orgs exposed. The bottleneck was never writing code. It was reviewing it. Nobody fixed that part. ⚠️ #DevSec #CyberSecurity
39
Nechiforel David Samuel
Nechiforel David Samuel @nsdhso ·
Security 101: Claude Code Access Control cd my-project; claude (scoped) --add-dir ./config (explicit) --add-dir / (DANGEROUS) .gitignore for secrets One bad command = entire filesystem exposed. Teach security first. #ClaudeCode #DevSec
15
Grafana
Grafana @grafana ·
🤝 Time to bridge the #DevSec divide. We've partnered with@MiggoSecurity on a joint solution that builds on production telemetry in Grafana Cloud to deliver evidence-based runtime security. grafana.com/blog/grafana-c…j
Bridge the DevSec divide: Using Grafana Cloud and Miggo for runtime protection | Grafana Labs

Grafana Labs and Miggo Security have partnered on a joint solution that builds directly on production telemetry in Grafana Cloud to deliver evidence-based runtime security.

From grafana.com
1
882
Salman Aslam
Salman Aslam @xalmanaxlam ·
AI now writes 30% of code at Microsoft and ~25% at Google. But studies show AI-generated code carries 2.74x more security vulnerabilities than human-written code. Speed without review is deferred risk. Your code gates matter more now, not less. #AI #DevSec #SoftwareEngineering
3
Gray Hats
Gray Hats @the_yellow_fall ·
Microsoft warns of a new campaign targeting engineers via fake Next.js technical assessments. Malware hides in VS Code tasks and npm scripts to steal secrets. #DevSec #NextJS #CyberSecurity #MicrosoftDefender #SoftwareEngineering #InfoSec #Malware securityonline.info/the-interview-…
The Interview Trap: Malicious Next.js Repositories Weaponize Coding Tests to Hack Developers

Microsoft warns of a new campaign targeting engineers via fake Next.js technical assessments. Malware hides in VS Code tasks and npm scripts to steal secrets.

From securityonline.info
237
LordePlayer
LordePlayer @LordePlayerX ·
People, be careful with these non-open source workflow automation projects. I tried @CompozyAI to test it, but before using it, I analyzed it to see where the data was going. Guess my surprise? 😲 #bigeye #bigbrother #devsec
1
1
126
Gabriel Abi Ramia → tubespark.ai
Gabriel Abi Ramia → tubespark.ai @gabrielabiramia ·
Devs gastando energia brigando por posts "limpos" enquanto entregam repositórios inteiros pro Claude/Cursor sem piscar. 849 upvotes reclamando de spam vs 59 discutindo se terceiros controlam quando vocês podem ser hackeados. Prioridades, né? #ClaudeCode #Cursor #DevSec
30
✪ 𝕱𝖆𝖍𝖆𝖉
✪ 𝕱𝖆𝖍𝖆𝖉 @fad_777 ·
برمجية Evelyn Stealer الخبيثة تستغل إضافات VS Code لسرقة بيانات اعتماد المطوّرين والعملات الرقمية #الأمن_السيبراني #Malware #DevSec thehackernews.com/2026/01/evelyn…
Evelyn Stealer Malware Abuses VS Code Extensions to Steal Developer Credentials and Crypto

Experts reveal Evelyn Stealer malware abusing VS Code extensions to steal developer credentials, browser data, and cryptocurrency wallets on Windows.

From thehackernews.com
37
yama | software developer 🇯🇵
yama | software developer 🇯🇵 @streamtech_en ·
Heads up: 900K+ users affected by fake Chrome extensions stealing ChatGPT/DeepSeek conversations. A reminder to audit your browser extensions periodically... 🔍 #CyberSecurity #DevSec stream-tech.dev/en/media/techn…q
Tech (General) Tech News (2026/01/09) | Stream Tech AI

Today's highlights focus on a significant structural shift in the software ecosystem: the decline of traditional developer communities and open-source business

From stream-tech.dev
101
Alexander Kazanski
Alexander Kazanski @alex_kazanski ·
Privacy in Copilot development: Your code isn't used to train models unless you opt-in. Microsoft prioritizes data security for enterprise users. #DevSec
13