Building safer AWS agents with Strands Decider 2B, Amazon Bedrock, and AgentCore Runtime to validate tool calls, prevent unsafe actions, and clarify ambiguity.

AccessLens AI combines Playwright, axe-core 4.13.0, and Strands Decider 2B to audit civic websites and prioritize accessibility findings. A policy gate uses confidence and severity thresholds to distinguish advisory remediation from human review, while Amazon Bedrock provides explanations without directly modifying public websites.

Encoding every patch of a gigapixel pathology slide with a foundation model is costly and redundant. PathFinder lets Strands Decider 2B choose, round by round, which patch receives the next H-optimus-0 inference and when to stop, encoding 16 of 12,590 patches in a real run. A reusable decision pattern for budget-aware agents.

Supply chains face sudden disruptions, from transit delays to cooling failures and cargo damage. CapacityX combines Strands Agents with Strands Decider 2B to select mitigation actions for logistics incidents using structured, confidence-aware decisions. This article explores the architecture, five operational actions, evaluation results, and lessons learned while building a smart logistics decision agent, including the challenges of integrating generative AI with reliable tool selection.

AgentX is a safety-focused AI agent prototype built with Strands Agents and Strands Decider. It evaluates proposed tool calls, distinguishes between proceed, ask first, and deny decisions, and records safety outcomes using simulated tools. The project demonstrates transparent, responsible AI-agent workflows without executing real-world actions.

Amazon’s open source Strands Decider 2B signals a shift from chatbot competition toward smaller AI systems that route tasks, select tools and control agent workflows efficiently.

Learn how I used Strands Decider 2B to route student requests in 2.4 seconds instead of 14.5—and how to apply this pattern to any decision-making problem. 100% accurate, brutally fast, and honest about uncertainty.

Strands Decider 2B checks every camera-trap frame before a vision model does. On cameras it had never seen, it cleared nearly half the empty frames for free and still caught 96.7% of animals.

A text adventure where the thing deciding what is true cannot write a word. I typed a bluff into my own game and it handed me the key - so I took the deciding away from the writer. Amazon Nova invents the story; Strands Decider 2B rules on it.

In a text RPG you can type anything, and almost none of it needs a large language model to decide what happens. I built a game master where a 2B Strands Decider model routes each action in one local request, plain Python enforces the rules, and Amazon Bedrock only writes the story.

Earth is too far, so Rovy decides: a Strands Decider 2B rover agent

I built AgentLoop on Amazon Bedrock with Strands Decider 2B to catch agents stuck in loops. Here’s what worked, what didn’t, and the honest numbers.

Discover how Hush uses Strands Decider 2B and Amazon Bedrock to create a zero-disturbance in-meeting AI wingman for Google Meet. Learn how confidence-aware restraint helps Hush alert users when goals are discussed, suggest timely doubts during pauses, and capture action items without clutter.

Parcel Detective is an AI-assisted warehouse investigation system that helps workers trace missing parcels using shipment evidence. It combines Strands Decider 2B for tool selection, Amazon Bedrock integration for grounded explanations, and a Python dashboard deployed on AWS EC2. Explore the architecture, investigation workflow, testing results, and lessons learned.

SmartTest AI classifies software bugs by severity, evaluates escalation needs, and recommends testing actions. Built with Strands Agents and Strands Decider 2B, it compares accuracy, latency, confidence, and cost against an LLM baseline, achieving 91.7% accuracy on 12 test cases with 5.9x lower median latency.

CyberShield AI is an AI-assisted security alert triage agent that helps analysts evaluate alerts using supplied evidence, risk-aware recommendations, and human-review guardrails powered by Strands Decider 2B.

SourceSignal AI is a research agent that uses Strands Decider 2B to decide when to retrieve more evidence, answer a question, investigate uncertainty, or abstain. This article explores its architecture, evaluation results, and deployment on Amazon Bedrock AgentCore.

Inbound lead gate for GTM advisory outreach. A Strands agent asks Decider 2B three typed questions about a founder or recruiter signal, then plain Python routes on the calibrated probability of outreach_worthy: PRIORITY (>=0.70), WARM (>=0.50), else SKIP. Only PRIORITY leads get a Nova Pro draft; failures fall back to WARM with human review and no draft. Runs on Bedrock AgentCore Runtime. In a live run of 12 leads, 3 got drafts. Reusable: swap the rubric, tune thresholds.

BugTriage AI: A Small Decision Model Decides, an LLM Explains

Building an AI-Assisted Incident Triage and Recovery Platform for Data Pipelines

Exploring how TM Forum ODA, operations research, infrastructure costs, and observability inform cross-architecture decisions through an experimental AI agent.

A hands-free kitchen helper for people with low vision: Strands Decider 2B decides who a phrase is for, whether it is about medicine, and which Amazon Nova model answers. Measured on 120 phrases I recorded and 52 label photos: 21 of 21 held-out medicine questions refused, 0.82 s to the first word, $0.11 of Bedrock per 1,000 phrases.

Every sculpture has a story. StoryStone investigates the clues and discovers it.

AWS has introduced two open-source projects for AI agent development: Strands Decider 2B, a small model designed to make decisions within agent workflows, and Strands Box, a sandbox for controlling agent execution. Both projects address different aspects of building agents that use tools, execute code, and interact with external systems.