Agent Native·1d agoClaude Bug HunterTo guard against cyber attacks, you have to package security methodologies, patterns from disclosed reports, tool integrations, and…
Sunaina Anumukonda·3d agoReplicating Lakera Gandalf: Testing AI Agent Vulnerabilities to Automated Prompt InjectionsTrack: Tech (AI Cybersecurity & Safety) Original Being Replicated: Lakera Gandalf AI Security Challenge Success Definition: Build an…
InAI SecuritybySanjeev Jaiswal (Jassi)·Oct 28The Non-Human Identity Crisis: Understanding Agent Identity with examplesHow NHI Governance Became the Unsolved Layer in Your Agent Security: and What We Can Build Now
Agent Native·Oct 23Black October: How To Defend Yourself Against Agentic Security HellOctober 2026 will be remembered as the month software security changed forever but much of what happened could have been prevented with…A response icon1A response icon1
Demetrio Milea·Oct 23Moving AHEAD: Rethinking SecOps Metrics for Machine Speed DefenseThis week has been quite a ride. In just a few days, defenders have had to assess two critical PHP Core vulnerabilities enabling…
Sandhya Kodali·Oct 21Responsible AI for Autonomous Agents: Where NVIDIA OpenShell Fits in the OWASP Agentic Top 10Responsible AI for autonomous agents isn’t achieved through prompt instructions or a one-time pre-deployment review. It requires a living…
InAI SecuritybySanjeev Jaiswal (Jassi)·Oct 21The only article you would ever need to understand Agentic AI Security: AI Agents Part 2When your agent becomes the attacker’s favourite target and why your defences are probably wrong.
Prophaze·Oct 21LLM Firewalling: 3 Real-World Security Findings That Show Why Prompt Filters Aren’t EnoughNo one wants to deploy an AI agent expecting that some customer request or webpage will inject code that will execute on your servers. But…
InSystem WeaknessbyMrĐức · AI & OT Security·Oct 16Build a Harmless Canary MCP Server to Test Your AI Agent for Tool PoisoningA 60-line Python lab that shows, with evidence, whether your agent obeys instructions hidden in tool metadata and tool output.
InOpen IntelligencebyDr. Fadi Shaar·Oct 14PentAGI: The Self-Hosted AI Team That Hacks Your Application Before Someone Else DoesHow a multi-agent system plans, executes, and remembers autonomous penetration tests inside an isolated sandbox
