MAY 12, 2026
EngBrief
Search⌘K
LatestTopicsSourcesSaved
Eng&Brief

Engineering insights from the world's best tech companies, curated and summarized.

Weekly brief

Browse

TopicsSourcesFavorites

More

SearchRSS Feed
© 2026 EngBriefUpdated every 4 hours
← Sources
kiro.dev icon
AI Agents

Kiro Blog

25 articles on EngBrief

Kiro is AWS's agentic IDE built for spec-driven development. The blog covers how the team is rethinking the software development workflow — from natural-language specs to production code — using AI agents, Bedrock models, and deep AWS integration to bring engineering rigor to agentic development.

AI AgentsDeveloper ToolsIDEAWS
Visit blog →

Latest Articles

Kiro3d ago

More room to explore: $20 paid tier sign-up bonus

Kiro has introduced a $20 sign-up bonus for new paid subscribers, doubling the previous credit limit and providing full model access from day one. This change aims to give developers sufficient runway to try Kiro before deciding on its suitability, with users getting access to premium models, including Claude Opus 4.7. The free tier remains unchanged, with free users having access to capable open weight models like Qwen3 Coder Next and DeepSeek v3.2.

AIDevTools
1 min
Kiro4d ago

Introducing Kiro Ambassadors

Kiro Ambassadors is a new program that selects engaged developers to collaborate closely with the Kiro team, providing feedback and influencing the product roadmap. In return, ambassadors receive a free Kiro subscription, early access to new features, and direct communication with the product and engineering teams. They commit to sharing their experience and product knowledge through content, events, and feature testing. The program aims to deepen the influence of developers who are active Kiro users, providing a platform for them to shape the product and drive meaningful technical extensions to the community. Ambassadors dedicate around 3-4 hours per month, including a monthly call with the Kiro engineering team and content creation or event participation.

AIDevTools
1 min
Kiro18d ago

Build with Kiro: Introducing the community hub and Kiro Labs

Kiro introduces the community hub, a centralized destination for Kiro developers to discover community projects, resources, and events. Kiro Labs, a new GitHub organization, showcases open-source projects built by Amazon employees to extend and enhance the Kiro development experience. The community is invited to participate by trying projects, filing issues, and sharing feedback, helping shape Kiro's evolution and drive innovation.

AIDevTools
1 min
Kiro25d ago

Opus 4.7 is now available in Kiro

Anthropic's Opus 4.7 model is now available in Kiro IDE and CLI, offering superior coding performance on complex tasks and resolving more production tasks with precise follow-through. This upgrade enables Kiro users to implement multi-step tasks across multiple files and tools with less drift and higher fidelity. Opus 4.7 is gradually rolling out to select Kiro customers with experimental support, with broader availability and features to follow.

AIDevTools
1 min
Kiro25d ago

Root cause in 33 seconds: How Kiro CLI saved 4 years of build time

Kiro CLI's root cause analysis toolchain identified a performance bottleneck in a legacy codebase, reducing build times from over 30 minutes to under 1 minute. The tool analyzed profiling data and code structure to pinpoint a config file parser consuming 80% of CPU time due to repeated re-initialization. Using a combination of code intelligence, file system, and subagent tools, Kiro implemented a fix by introducing a shared cache, reducing 4 years of monthly compute time and accelerating future investigations by leveraging AI-assisted development capabilities.

AIDevTools
1 min
Kiro28d ago

Run Kiro CLI programmatically: introducing headless mode

To run Kiro CLI programmatically, a headless mode was introduced, allowing for automated workflows without a browser. This is achieved by generating an API key and setting it as an environment variable, KIRO_API_KEY, which skips the browser-based login flow. A concrete example demonstrates an automated code reviewer that runs on every push via GitHub Actions, utilizing a custom Kiro agent and a GitHub Actions workflow to review code for security issues and best practices.

AIDevTools
1 min
Kiro28d ago

Kiro CLI 2.0: a new look and feel, headless CI/CD pipelines, and Windows support

Here is a 3-sentence summary of the Engineering Blog Post: Kiro CLI 2.0 introduces three major upgrades: headless CI/CD pipelines, native Windows support, and a refreshed user experience (UX). Headless mode enables developers to run Kiro CLI programmatically within their CI/CD pipelines, automating deployments end-to-end. The new UX features a subagent experience, task list, and improved monitoring capabilities, allowing developers to build complex codebases, automate workflows, and analyze errors more efficiently.

AIDevTools
1 min
KiroApr 10, 2026

Planview saves 40+ hours per audit cycle by automating SOC 2 compliance with Kiro CLI

Planview automated SOC 2 compliance by creating a custom agent in Kiro CLI, saving over 40 hours per audit cycle. The agent provided read-only access to AWS services, automating evidence collection and reducing manual labor for engineers. With Kiro's help, Planview simplified the SOC 2 auditing process, allowing them to pull information, verify settings, and generate compliance evidence with timestamps.

AIDevTools
1 min
KiroApr 7, 2026

We’re bringing back the Kiro startup credits program

Kiro is reviving its startup credits program, offering up to a year of free access to Kiro Pro+ to eligible early-stage startups. The program provides three tiers based on team size, allowing startups to scale with Kiro's specs and AI agents without worrying about costs. This decision aims to provide structure to AI coding workflows, increase code quality, and minimize development time, enabling founders to achieve sustainable progress and scale their teams.

AIDevTools
1 min
KiroApr 2, 2026

MiniMax M2.5 and GLM-5 are now in Kiro

Kiro now supports MiniMax M2.5 and GLM-5, two open-weight models that extend the IDE's modeling capabilities. MiniMax M2.5 offers high efficiency and performance for complex tasks, while GLM-5 excels at long-horizon agentic workflows and large codebases. Both models are directly accessible through the Kiro IDE and CLI, with experimental support available for various regions.

AIDevTools
1 min
KiroMar 26, 2026

Bringing agentic AI to silicon development

Here is a 2-3 sentence summary of the blog post: Annapurna Labs' agentic AI architecture bridges the gap between general-purpose LLMs and hardware development by integrating with proprietary EDA tools and complex infrastructure. This approach provides an AI agent, like Kiro, with the ability to interact with hardware development workflows, generating high-quality code and providing rapid feedback. The real-world impact of this technology includes a 6x speedup in testbench development and transformative results for hardware teams.

AIDevTools
1 min
KiroMar 24, 2026

From copilots to coworkers at AAAI: the gap between agentic research and production

Here is a 3-sentence summary of the engineering decisions and outcomes from various companies attempting to deploy coding agents into production: Research and production have fundamentally different optimization goals, with research prioritizing capability and production focusing on reliability, cost, latency, trust, and organizational fit. Companies like Microsoft, LinkedIn, and Mistral have encountered challenges in deploying coding agents, including architectural, orchestration, and evaluation complexities, which require a system-level design approach that balances cost, latency, and quality. Ultimately, developing and evaluating coding agents in production demands an engineering-focused approach that prioritizes scalable RL environments, robust evaluation benchmarks, and infrastructure to support agent training and evaluation at scale.

AIDevTools
1 min
KiroMar 23, 2026

A new look for the Kiro CLI

Kiro CLI has been updated with a refreshed user experience, featuring a live unified status bar, omnipresent input area, and contextual overlays. The new design aims to reduce friction and improve interaction with the Kiro agent, while maintaining the natural flow and directness of a terminal environment. Key features include a status bar that displays active and previous sessions, an input area that remains available at all times, and customizable context overlays for managing settings and credits.

AIDevTools
1 min
KiroMar 18, 2026

Introducing the Kiro Students tier

Kiro introduces the Kiro Students tier, offering one year of free access with 1,000 credits per month to eligible university students, without credit card or trial timer requirements. The tier aims to provide students with professional-grade tools to build, experiment, and learn without pressure, leveraging AI agents for development. With Kiro, students can decompose complex problems into manageable steps, build structured thinking, and move seamlessly between planning and shipping code.

AIDevTools
1 min
KiroMar 12, 2026

Enterprise governance: control your MCP servers and models

Kiro introduces two key features to enhance enterprise governance: an MCP server registry and model governance controls. The MCP server registry allows administrators to define and enforce which MCP servers are approved for use, ensuring centralized oversight and adherence to organizational policies. Administrators can host a JSON registry file on any HTTPS endpoint and configure Kiro to fetch and re-sync the registry regularly. Model governance enables organizations to restrict which AI models are available to their developers, reflecting the need for compliance review before model usage. Administrators can disable unapproved models, preventing developers from accessing them, and manage the availability of models based on their data residency requirements and compliance obligations.

AIDevTools
1 min
KiroFeb 27, 2026

Surgical precision with AST-based code editing in Kiro

Kiro's new AST-based code navigation and editing engine reduces token usage by 20% and enables precise, resilient code transformations, improving performance and reducing errors. The engine targets code by structure, applying typed operations with semantic understanding, unlike traditional text-based approaches that burn through tokens and break easily. By using abstract syntax trees, Kiro's editor provides 58% fewer tokens and 73% fewer tokens for specific code operations, resulting in improved efficiency and maintainability.

AIDevTools
1 min
KiroFeb 23, 2026

The hidden inefficiencies in AI coding (and how we find them)

Here's a 2-3 sentence summary of the article "The hidden inefficiencies in AI coding (and how we find them)" by Kiro: To improve AI coding agent efficiency, a specialized system called CORAL was built for continual optimization via reasoning and adaptive learning. The system analyzes real agent interactions to surface patterns that benchmarks and pass/fail metrics miss, such as silent search failures and incorrect shell command usage, and extracts generalizable lessons to improve workflow patterns and error recovery. By automatically identifying and addressing inefficiencies, CORAL enables the development of high-confidence insights that can be turned into concrete fixes, reducing incorrect pattern rates and improving overall agent performance.

AIDevTools
1 min
KiroFeb 19, 2026

The bug fix paradox: why AI agents keep breaking working code

Here's a 2-3 sentence summary of the technical content: Kiro's property-aware code evolution methodology helps define explicit boundaries between bugs to fix and behaviors to preserve by creating a bug condition (C) that partitions the input space, a postcondition (P) that defines what "fixed" means, and fix and preservation properties that constrain how AI agents write fixes. AI agents like Kiro then use these properties to propose a bugfix workflow, which includes formulating a root cause hypothesis, testing it through specific tests, and writing a fix that must pass both the fix and preservation properties. This approach prevents AI agents from making unnecessary changes and ensures that the fix respects the original intent, reducing the likelihood of the agent introducing new bugs.

AIDevTools
1 min
KiroFeb 18, 2026

New spec types: fix bugs and build on top of existing apps

**Design-first Specs:** Kiro's new Design-first workflow lets developers start with a technical approach, matching their existing thought process when building on existing apps or fixing bugs. By choosing a high-level or low-level design, developers can explore different technical approaches and ensure requirements are feasible. This workflow is ideal for projects with pre-defined architectures, prototyping, or strict non-functional requirements. **Bugfix Specs:** Kiro's Bugfix workflow helps developers fix complex bugs with precision by defining the scenario that caused the error, stating what should not be modified, and generating a structured document with Current Behavior, Expected Behavior, and Unchanged Behavior sections. This ensures accurate testing and minimizes regression risk, making it suitable for complex bugs, high-risk changes, or documentation-heavy fixes.

AIDevTools
1 min
KiroFeb 18, 2026

Kiro loves regulated workloads

Kiro, a developer tool, is now available in the AWS GovCloud (US) Regions, enabling government developers to access its AI-driven capabilities within their regulated environments. This launch supports private connectivity for restricted networks and provides features for security and compliance, including encryption and enterprise authentication. Kiro's deployment in the AWS GovCloud (US) Regions meets elevated compliance needs and offers a modern dev experience without compromising on capability or introducing workarounds for government-specific constraints.

AIDevTools
1 min