DEV Community

← All Trends

AI Agent Security & Blast Radius Control

24 posts in this trend in the last 7 days • Active about 5 hours ago
Your AI Agent Will Do Something Terrible. Here's How to Survive It.

Painful lessons from early deployers
james_anderson_h profile

Your AI Agent Will Do Something Terrible. Here's How to Survive It.

#ai #agents #security #webdev
28
Comments 23
7 min read
AI Agent Breach 101
maltsev-dev profile

AI Agent Breach 101

#agents #ai #architecture #security
Comments
2 min read
Your AI agent doesn't need fewer permissions. It needs a mission.
chanpre75813933 profile

Your AI agent doesn't need fewer permissions. It needs a mission.

#ai #security #agents #webdev
Comments
2 min read
Is "more human oversight" actually the right fix for AI agent incidents?
caiderek profile

Is "more human oversight" actually the right fix for AI agent incidents?

#discuss #ai #security #programming
1
Comments
3 min read
AI agent security risks: 4 controls for 2026
azrty profile
Azrty

AI agent security risks: 4 controls for 2026

#aiagents #security #promptinjection #mcp
Comments 1
12 min read
The Wiki Write That Wasn't a Write: Why Agent Guardrails Must Check Outcomes, Not Tools
syncsoftai profile

The Wiki Write That Wasn't a Write: Why Agent Guardrails Must Check Outcomes, Not Tools

#ai #security #machinelearning #agents
Comments 1
6 min read
A Coding Agent's Permissions Need a Negative Test
hefty_69a4c2d631c9dd70724 profile
hefty

A Coding Agent's Permissions Need a Negative Test

#ai #devtools #security #testing
Comments
4 min read
We Gave AI Agents Real Tools — Then Realized “Just Ask Before Acting” Wasn’t Enough
robertadam987_ profile

We Gave AI Agents Real Tools — Then Realized “Just Ask Before Acting” Wasn’t Enough

#ai #security #programming #softwaredevelopment
14
Comments 14
6 min read
Always-On AI Agents Need a Permission Architecture, Not Just Better Prompts
shizhe_lim_61b5948cb8081d profile

Always-On AI Agents Need a Permission Architecture, Not Just Better Prompts

#ai #security #architecture #agents
Comments 1
3 min read
Plan Mode vs. Auto-Accept: How Execution Mode Selection Becomes a Safety Primitive in Coding Agents
mech_app_ai profile

Plan Mode vs. Auto-Accept: How Execution Mode Selection Becomes a Safety Primitive in Coding Agents

#agents #ai #security #softwaredevelopment
1
Comments
6 min read
Prompt Injection Is Coming for Your Coding Agent (And a README Can Be the Attack)
codenamew profile

Prompt Injection Is Coming for Your Coding Agent (And a README Can Be the Attack)

#ai #security #devops #opensource
1
Comments 2
3 min read
Why Prompts Fail as AI Agent Guardrails (And How to Fix It)
rcortez056 profile

Why Prompts Fail as AI Agent Guardrails (And How to Fix It)

#ai #security #systemdesign #architecture
Comments
2 min read
Your repo is not trusted context. What I changed after giving coding agents real repositories
bloqarl profile

Your repo is not trusted context. What I changed after giving coding agents real repositories

#security #ai #devops #programming
2
Comments 2
3 min read
I Tried to Make My Own AI Agent Leak Its Secrets (and It Did)
seek3r profile

I Tried to Make My Own AI Agent Leak Its Secrets (and It Did)

#ai #security #llm #webdev
Comments
4 min read
Your Agent Should Not Borrow Your API Key
anusha_mukka profile

Your Agent Should Not Borrow Your API Key

#ai #security #agents #iam
2
Comments 1
11 min read
A CVE About AI Agent Permissions Made Me Re-Check What Our Own Agent Could Touch
rbonweb profile

A CVE About AI Agent Permissions Made Me Re-Check What Our Own Agent Could Touch

#ai #security #devops #programming
4
Comments
5 min read
AI Should Propose. Systems Should Verify.
sinarezaei profile

AI Should Propose. Systems Should Verify.

#ai #security #devops #architecture
4
Comments 5
4 min read
Human in the loop for AI agents: which actions need approval and which don't
forrestzhang profile

Human in the loop for AI agents: which actions need approval and which don't

#ai #agents #security #llm
2
Comments 1
11 min read