AWS wellarchitected: Add AGENTREL02 task-execution best-practice links
Summary
Adds five links to AGENTREL02 best-practice pages, including limiting agent permissions to minimum required access and behavioral anomaly detection.
Security assessment
One added link documents least-privilege permission scoping for agents, which is security guidance, but the diff is only a link list and does not describe or remediate a specific vulnerability.
Evidence
+[AGENTREL02-BP02 Limit agent permissions to minimum required access](./agentrel02-bp02.html)
Diff
diff --git a/wellarchitected/latest/agentic-ai-lens/agentrel02.md b/wellarchitected/latest/agentic-ai-lens/agentrel02.md index a1e7eb1..f39f4b2 100644 --- a//wellarchitected/latest/agentic-ai-lens/agentrel02.md +++ b//wellarchitected/latest/agentic-ai-lens/agentrel02.md @@ -14,0 +15,5 @@ AGENTREL02: How do you develop agentic systems that reliably execute tasks with +[AGENTREL02-BP01 Design agents for specific and atomic tasks](./agentrel02-bp01.html) +[AGENTREL02-BP02 Limit agent permissions to minimum required access](./agentrel02-bp02.html) +[AGENTREL02-BP03 Implement behavioral anomaly detection and monitoring](./agentrel02-bp03.html) +[AGENTREL02-BP04 Develop clear instruction protocols for agents](./agentrel02-bp04.html) +[AGENTREL02-BP05 Establish tiered human oversight and approval workflows](./agentrel02-bp05.html)