Agent Kernel architectural principles, core abstractions, and design patterns. Use this skill when you need to understand the codebase structure, how components interact, or before making changes to core functionality. Covers Session, Agent, Runner, Module, Runtime, AgentService, AKConfig, tools, hooks, multimodal, conversation threads, the adapter pattern, and the AWS ECS containerized deployment classes (ECSIOHandler, ECSOutputConsumer, ECSAgentRunner, ECSSQSConsumer, QueueConsumer, ThreadRunner).
Step-by-step guide for adding a new multimodal attachment storage backend to Agent Kernel. Use this skill when you need to integrate a new storage service (beyond in-memory, Redis, and DynamoDB) for persisting image and file attachments. Covers implementing the AttachmentStore interface, factory registration, configuration, and testing.
Testing conventions, patterns, and automation for Agent Kernel development. Use this skill when writing tests for new features, debugging test failures, or understanding the test infrastructure. Covers pytest patterns, async testing, mocking, the built-in Test framework, and CI/CD test workflows.
Review a GitHub pull request against Agent Kernel's architecture, design principles, code quality standards, and testing conventions, then post the findings back to the PR as review comments. When the PR contains spec documents (design.md, spec.md, or plan.md under docs/specs/), they are reviewed first — each at its own stage's altitude — and any implementation is checked against them. Use this skill when given a PR number or URL to review, e.g. "review PR 342" or "run a review on https://github.com/yaalalabs/agent-kernel/pull/342".
Write the spec documents for a planned Agent Kernel change under docs/specs/<ticket>/ in three ordered stages: a concise point-form design spec (design.md) that a maintainer reviews first, then a detailed implementation spec (spec.md) once the design is approved, then a concise implementation plan (plan.md). Use this skill when asked to write or update a design, spec, or plan for a feature, refactor, or fix before it is implemented, e.g. "write a design spec for AK-52" or "spec out the shared driver extraction".
Add capabilities to an existing Agent Kernel project. This skill guides you through adding guardrails, tracing/observability, session persistence, knowledge bases, MCP server, A2A server, pre/post hooks, multimodal support, and conversation thread support. Session persistence supports Redis, DynamoDB (AWS), Cosmos DB (Azure), and Firestore (GCP). Conversation threads support in-memory, Redis, DynamoDB (AWS), Firestore (GCP), and Cosmos DB (Azure) backends. Generates configuration and code changes needed.
Deploy an Agent Kernel project to AWS, Azure, or GCP using Terraform modules. Supports serverless and containerized modes for all three clouds. AWS supports execution modes (rest_sync, rest_async, async, stream), queue-based scalable processing, and custom API Gateway authorizers. GCP supports Cloud Run serverless (scale-to-zero) and containerized (always-on) with Redis or Firestore session backends.
Code quality standards, formatting, commit conventions, and PR workflow for Agent Kernel development. Use this skill when making contributions, formatting code, writing commit messages, or preparing pull requests.