65
likes
9
comments
daily view 0
monthly view 0
Live Analytics
Comments 0
I Built an AI Security Gate. It Approved a Leak. Analytics Table
Income Estimates for I Built an AI Security Gate. It Approved a Leak.
Based on this YouTube video's total view count of 91.9K views and industry-standard rates, the estimated total earning is $64 - $184 through ad revenue. Historical data is not yet available to calculate daily, weekly, or monthly averages.
About I Built an AI Security Gate. It Approved a Leak.
Explore I Built an AI Security Gate. It Approved a Leak. with 91,943 views, 65 likes, and 9 comments. Experience the impact of this video content that has captured audience attention.
I built an AI security gate for Claude Code using Jev by TypeSafeâand then I attacked it. The surprising result: Jev correctly identified a payment-config upload as network egress and detected the note trying to steer the reviewer. But my own policy still returned ALLOW. The OS sandboxânot my AI gateâstopped the upload. In this engineering tutorial, I show the complete build: - Claude Code PreToolUse hooks - Jev typed questions and confidence scores - Allow, ask, and deny policy rules - Prompt-injection and reviewer-steering detection - Network-egress checks - Shell write-target validation - Native OS sandboxing - Confidence testing and model calibration - The final fix that changed the same attack from ALLOW to DENY - This is not a perfect âAI fixes securityâ story. It is a real engineering story about building safer AI coding agents. The model is a sensor. Your code is the judge. A confidence score can help veto an action, but it should never approve one by itself. A correct classification does not automatically make the underlying action safeâand every AI gate still needs an independent sandbox beneath it. CHAPTERS 00:00 The AI Gate Approved a Leak 01:03 A Model That Doesnât Talk 03:22 Building the Gate 05:07 Real Work 06:14 Breaking It 07:48 Every Layer Leaks 08:52 Fixing the Judge 10:55 Can You Trust 0.99? 13:01 The Rules TOPICS COVERED Claude Code hooks, AI coding agents, AI agent security, Jev, TypeSafe, prompt injection, LLM security, tool calling, OS sandboxing, network egress, confidence scores, model calibration, and secure agent workflows. #ClaudeCode #AISecurity #AIAgents
About YouTube Real-Time View Count
With SocialCounts.orgâs view counter, track your YouTube videoâs live view count and YouTube likes count in real time with fast, reliable updates.
Watch every YouTube video live view count rise with our real-time YouTube views trackerâbuilt for accuracy and minimal delay.
Follow YouTube real time views as they happen, using our dedicated view counter for YouTube videos.
Get up-to-date live view count on YouTube and see real-time growth with SocialCounts.orgâs smart tracking tools.
Embed Widget
Parameters:
fullscreen=true- Fullscreen countergraph=true- Live graph chartcounter=0/1/2- Select counter (0=likes, 1=views, 2=comments)
URL
Click to copy the embed URL to your clipboard

