DOCSEARCHbyAlgolia
Submit Docs

Kiro Powers

repository·main·Indexed 43 days ago

https://github.com/kirodotdev/powers

A collection of specialized toolsets and MCP (Model Context Protocol) servers designed to extend AI agent capabilities. It provides knowledge bases and workflows for cloud infrastructure, observability, and platform-specific development, including a Databricks Power and an Arm SoC Migration Power for architecture analysis and cross-compilation.

Tokens
372.4K
Snippets
719
Records
1.3K
Agent score
56%

What's inside Kiro Powers

  1. Overview of AWS Observability capabilities

    main

    The aws-observability power provides a unified interface for:

    • CloudWatch Logs: Query and analyze logs via CloudWatch Logs Insights.
    • Metrics & Alarms: Query metrics with Metrics Insights and receive alarm recommendations.
    • Application Signals: APM features including distributed tracing, service maps, and SLOs.
    • Amazon Managed Prometheus (AMP): Execute PromQL queries against AMP workspaces using SigV4 authentication.
    • Codebase Observability Analysis: Automated identification of missing instrumentation or logging gaps in your code.
    • CloudTrail Integration: Security auditing and tracking API activity/IAM changes.
    • AWS Documentation: Direct access to official AWS documentation for troubleshooting.

  2. Overview of StackGen Power

    main

    StackGen (Aiden for InfraOps) is a Kiro power designed for designing and managing cloud infrastructure as code (IaC). It allows you to build, operate, observe, and remediate infrastructure across AWS, Azure, and GCP.

    Key capabilities include:

    • Appstack Management: Creating and managing infrastructure configurations.
    • Resource Operations: Adding, updating, and connecting cloud resources (EC2, S3, VPC, etc.).
    • Environment Management: Configuring environment-specific profiles (dev, staging, prod).
    • Git Integration: Pushing IaC to Git repositories for version control.
    • Policy & Compliance: Scanning configurations against benchmarks and detecting violations.
    • Deployment & Operations: Triggering Plan/Apply/Destroy runs via governed SaaS runners and detecting configuration drift.

  3. Overview of Cloud Architect Power

    main
    Cloud Architect Power is a specialized toolset designed to assist in building AWS infrastructure using the AWS CDK with Python. It enforces the AWS Well-Architected framework and provides AI-driven assistance for infrastructure design, implementation, and optimization. It integrates multiple MCP (Model Context Protocol) servers to provide real-time AWS pricing, API execution, and documentation access.

  4. Overview of the GCP to AWS Migration Advisor

    main

    The migration-to-aws power (GCP to AWS Migration Advisor) is designed to migrate workloads from Google Cloud Platform to AWS. This includes standard infrastructure and specialized AI/agentic workloads.

    Supported Migration Triggers

    The advisor triggers on various migration intents, such as:

    • Moving from GCP to AWS (e.g., Cloud Run → Fargate, Cloud SQL → RDS, GKE → EKS).
    • Migrating Terraform IaC to AWS.
    • Migrating AI workloads (e.g., OpenAI/Gemini/LangChain to Amazon Bedrock).

    Migration Process

    The advisor follows a 6-phase process:

    1. Discover: Identify GCP resources from Terraform files, app code, or billing exports.
    2. Clarify: Define migration requirements (must be completed before Design, Estimate, or Generate).
    3. Design: Architect the AWS equivalent.
    4. Estimate: Calculate costs.
    5. Generate: Produce migration artifacts.
    6. Feedback: Collect optional user feedback.

    Key Philosophies

    • Re-platform by default: Matches GCP service types to AWS equivalents (e.g., Cloud Run to Fargate).
    • Dev sizing by default: Uses development-tier capacity (e.g., db.t4g.micro) unless specified otherwise.
    • No human labor costs: Estimates focus on vendor charges (like data transfer egress) rather than professional services or people-time.
    • BigQuery Handling: The advisor does not recommend a specific AWS analytics service for BigQuery. Instead, it triggers a specialist advisory gate, directing users to their AWS account team or a data analytics partner.

  5. Overview of Dynatrace Observability Power

    main

    The Dynatrace Observability Power provides access to Dynatrace monitoring data using Dynatrace Query Language (DQL). It allows for querying logs, metrics, distributed traces, problems, and Kubernetes events, and leverages Davis AI for intelligent insights.

    Key Capabilities:

    • DQL Querying: Execute queries against the GRAIL data lake.
    • Natural Language to DQL: Automatically convert questions into DQL queries.
    • Problem Management: Track and analyze active/closed problems via Davis AI.
    • Entity & Topology: Locate services, hosts, processes, and their relationships.
    • Kubernetes Monitoring: Query cluster events, pod conditions, and resource usage.
    • Security: List vulnerabilities and access security assessments.
    • Notifications: Send alerts via email or Slack.
    • RUM: Analyze user sessions and Core Web Vitals.

    Authentication Requirements:

    • Dynatrace environment URL
    • Bearer token

  6. Overview of Databricks AI Dev Kit Power

    main

    The Databricks AI Dev Kit Power provides comprehensive access to the Databricks Lakehouse platform. It uses expert steering files to guide development across various domains like pipelines, jobs, dashboards, ML, governance, and apps.

    Additionally, it offers an optional MCP server that provides 44 live tools (over 180 operations). Note that the steering files work independently of the MCP server. The MCP server is installed separately and is disabled by default in the mcp.json configuration until manually enabled.

    Key Capabilities:

    • SQL & Compute: Execute SQL on warehouses, manage cluster lifecycles, and run Python/Scala.
    • Pipelines & Jobs: Build Spark Declarative Pipelines (streaming tables, CDC, Auto Loader) and orchestrate multi-task DAGs.
    • Unity Catalog: Manage tables, volumes, grants, storage credentials, and External Iceberg Reads.
    • AI/ML: Vector Search, Model Serving (including FMAPI), MLflow tracking/evaluation, and Agent Bricks (RAG/Supervisors).
    • App Development: Build full-stack web applications on the Lakehouse and provision Lakebase (managed PostgreSQL).
    • Infrastructure: Use Databricks Asset Bundles for IaC.

    Authentication Requirements: Requires a Databricks workspace and one of the following:

    • An OAuth-authenticated CLI profile.
    • An existing ~/.databrickscfg profile.
    • A Personal Access Token (PAT).

  7. Overview of Neon Database Power

    main

    The Neon power provides complete access to Neon's MCP server for managing serverless Postgres. It allows for project management, database branching, schema management via SQL execution, and retrieving connection strings.

    Key Features:

    • Database Branching: Create isolated copies of your database for development or testing.
    • Project Management: Create and manage Neon projects and organizations.
    • Schema Management: Execute SQL queries to manage tables and schemas.
    • Connection Strings: Retrieve PostgreSQL connection strings for specific projects or branches.

    Authentication: No API keys are required; authentication is handled via the Neon CLI or browser login.

  8. Overview of VMware Network Migration capabilities

    main

    The VMware Network Migration capability allows you to migrate on-premises network infrastructure to AWS. It translates source environment configurations into AWS-equivalent resources including VPCs, subnets, security groups, NAT gateways, transit gateways, elastic IPs, routes, and route tables.

    Key Features:

    • Resource Translation: Converts network segments to VPCs/subnets and firewall rules to Security Groups.
    • Routing Options: Supports AWS Transit Gateway (Hub and Spoke) or isolated VPC route tables.
    • VPC/Subnet Management: Supports renaming, resizing, merging, splitting, deleting, and changing IP addresses.
    • IaC Generation: Produces CloudFormation, CDK (TypeScript), Terraform, or Landing Zone Accelerator (LZA) code.
    • Visualization: Generates network diagrams in PNG and Mermaid formats.
    • Analysis: Enables reachability analysis on deployed VPCs.

  9. Overview of VMware Migration capabilities

    main

    AWS Transform enables the migration of VMware environments to AWS using generative AI for planning and execution. The process orchestrates the full lifecycle from discovery to server rehosting on Amazon EC2.

    Key Capabilities:

    • VMware VMs → Amazon EC2: Rehost/lift-and-shift via AWS Application Migration Service (MGN).
    • AI-Driven Network Conversion: Translates VMware network configurations into AWS VPC architecture (VPCs, subnets, security groups, Transit Gateway).
    • AI-Driven Migration Planning: Generates application grouping, wave planning, and right-sizing recommendations.
    • Discovery Options: Supports AWS Application Discovery Service (ADS) collectors, Export for vCenter tool, or independent data imports.
    • Landing Zone Setup: Automates AWS account structure, IAM roles, and baseline infrastructure.
    • Multi-Wave Migration: Supports complex migrations with per-wave configurations and multi-account support.
    • Containerization (Standalone): AI-driven generation of Docker artifacts, ECR publishing, and EKS/ECS IaC deployment.

  10. Overview of Kiro Powers

    main

    Kiro powers are specialized toolsets and workflows designed to enhance AI agent capabilities. Each power provides specific tools for development tasks, such as cloud infrastructure management, migrations, observability, and platform-specific integrations.

    While this repository contains a subset of available powers, the complete registry of all published powers can be found at kiro.dev/powers. Detailed documentation for powers is available at https://kiro.dev/docs/powers/.

  11. Overview of AWS Transform custom

    main

    AWS Transform custom is a tool for performing code upgrades, migrations, and transformations. It supports any-to-any transformations including language version upgrades (Java, Python, Node.js), framework migrations, AWS SDK migrations (e.g., Java SDK v1→v2, boto2→boto3, JS SDK v2→v3), library upgrades, and custom organization-specific refactoring.

    There are two execution modes:

    • Local mode: Runs the AWS Transform CLI directly on your machine. Best for 1-9 repositories (max 3 concurrent).
    • Remote mode: Runs transformations at scale via AWS Batch/Fargate containers. Best for 10+ repositories or when cloud execution is preferred. Infrastructure is auto-deployed with user consent.

  12. Use kiro-support to manage AWS Support cases

    main

    The kiro-support power allows you to file, track, reply to, and resolve AWS Support cases directly from your IDE. It uses two possible backends:

    1. Primary: awslabs.aws-support-mcp-server (provides typed tools with snake_case arguments).
    2. Fallback: aws-mcp (Agent Toolkit, runs aws support ... CLI commands with kebab-case flags).

    Both backends require a Business+ support plan and support:* IAM permissions.

    Common Workflows:

    • Raise a case: Tell the assistant to file a case (e.g., file a normal case: Kiro chat keeps disconnecting on macOS). The assistant will draft the body, show it to you for approval, and then file it.
    • Track/Read/Reply/Resolve: Use natural language to manage existing cases.
    • Check access: Ask to check access or setup to run a read-only probe.