Not every project starts from requirements, and now specs don’t have to either. The new design-first flow lets you start with your technical approach and work backward to requirements. Bugfix specs locks down exactly what should change and what shouldn’t before writing any code. Download or restart Kiro to try it. 👉 http://spr.ly/6045hQPCL
Kiro
Technology, Information and Internet
Agentic AI development from prototype to production.
About us
Kiro helps you do your best work by bringing structure to AI coding with spec-driven development.
- Website
-
https://kiro.dev
External link for Kiro
- Industry
- Technology, Information and Internet
- Company size
- 10,001+ employees
Updates
-
David Yanacek is a Senior Principal Engineer and a lead advisor on the Agentic AI team at Amazon Web Services (AWS). he joins Kevin Ball to discuss Kiro, Bedrock AgentCore, spec driven development, frontier agents, and the future of agentic software development. Listen here: https://lnkd.in/g5Rw5ZFg
-
Ever load a large codebase into your AI and watch it slowly lose the plot? That’s your context window hitting 90%. One agent, too much to hold, confused output. Sub-agents fix this ✅ · Database · Frontend components · Test files Each gets a clean context window. Each runs in parallel. Focused results. Try sub-agents in Kiro and see the difference yourself. 👉 http://spr.ly/6048htoyw
-
Looking to bring your Figma designs into Kiro? Join us as Akbar Mirza, Figma, and Nikhil Swaminathan, Kiro, show how to streamline design-to-development workflows across your favorite tools—mapping code components, implementing designs, and generating architecture diagrams with the Figma MCP server.
Kiro Bi-Weekly Show | Design to Code with Figma
www.linkedin.com
-
Calling all designers and devs 📣 We’re going live today at 9:30 am PT with Figma to show you how the Figma power streamlines your design-to-development workflows. From mapping code components to generating architecture diagrams. Less back-and-forth. More shipping. You won't want to miss it 👉 http://spr.ly/6048hUOfE
-
-
In case you missed it, Specs just got a lot more flexible. Two new Spec types shipped last week. Design-first for when you already have the architecture mapped out and just need help implementing it. Bugfix for when you know what’s broken and need a surgical fix without touching anything else. Download or restart Kiro to try it now 👉 http://spr.ly/6044hUDZa
-
-
Designers and devs, this one’s for you 🎨 We’re going live Wednesday at 9:30 am PT with Figma to show you how the Figma power streamlines your design-to-development workflows. From mapping code components to generating architecture diagrams. Less back-and-forth. More shipping. You won't want to miss it 👉 http://spr.ly/6047hm7oz
kirodotdev - Twitch
twitch.tv
-
Most bug reports describe what broke, not what must stay the same. That gap is where AI agents cause regressions. They fix the reported issue and quietly break something adjacent. Kiro solves this by making three things explicit before a single line of code gets written: what's broken, what should happen instead, and what must not change. The fix stays surgical because the boundary is defined upfront. Read how it works → http://spr.ly/6045hoFNl
-
-
With Great Power Comes Great Responsibility: How Datadog + Kiro powers Drive Responsible AI Development "With power comes great responsibility" – a quote that resonates deeply as we navigate the AI revolution. As developers gain unprecedented capabilities through AI agents and autonomous systems, the need for responsible, observable, and accountable AI has never been more critical. The Challenge: AI agents are powerful, but without proper observability, they become black boxes. How do we ensure our AI systems are performing as intended? How do we detect bias, monitor resource consumption, or troubleshoot failures in real-time? Enter Datadog + Kiro powers : By integrating Datadog's comprehensive observability platform with Amazon Kiro's autonomous agent capabilities, we're empowering developers to build AI systems that are not just powerful, but responsible. What This Means: ✅ Transparency: Real-time visibility into AI agent behavior, decisions, and performance metrics ✅ Accountability: Complete audit trails of AI actions with distributed tracing across microservices ✅ Reliability: Proactive monitoring to detect anomalies, bias, and performance degradation before they impact users ✅ Efficiency: Optimize AI workloads with insights into resource consumption and cost management ✅ Security: Monitor for malicious activity and ensure AI systems operate within defined guardrails The Integration: Through Model Context Protocol (MCP), Kiro agents can now access live production insights from Datadog directly within the development environment. Developers can query observability data, analyze system behavior, and make informed decisions – all through natural language. Real Impact: This isn't just about monitoring – it's about building AI systems we can trust. Systems that are explainable, auditable, and aligned with responsible AI principles. Systems that respect user privacy, maintain fairness, and operate transparently. As we continue pushing the boundaries of what AI can do, let's remember: the greatest innovations come not from raw power alone, but from wielding that power responsibly. The future of AI is observable. The future of AI is responsible. Bharadwaj Tanikella Kunal Batra Cole DeFreitas George Morgan Courtney Willison Arundeep Nagaraj Sujatha Kuppuraju Nina C.Varun Jasti Krishna Dalal Kolawole Akinnibi Dhilip venkatesh Uvarajan Andrew Hansen Derek Francis John McElhone Ross Warren Matt Auerbach Jonathan Morin Dan Kiuna Kiro Amazon Web Services (AWS) Datadog #datadog #kiro #devops #awspartners #mcp
-
Kiro reposted this
We just shipped Kiro CLI v1.26.0. My two favorite updates are 1. Automatic skill loading — Skills from .kiro/skills/ and ~/.kiro/skills/ are now automatically available to the default agent. Drop a skill in the directory and it's ready to go, no manual configuration needed. 2. Inline file references — Use @path syntax to inject file contents or directory trees directly into your messages. No tool calls, fewer tokens, faster context loading. Check out the full release notes here. https://lnkd.in/gqnM--ZX Next launch will be even better ;) What do you want to see us add next?