Addy Osmani drops Agent Skills for senior AI engineering

// 45d agoOPENSOURCE RELEASE

Addy Osmani drops Agent Skills for senior AI engineering

Google Engineering Lead Addy Osmani releases an open-source framework of 22+ production-ready workflows that enforce senior engineering discipline in AI coding agents. The project introduces quality gates and "anti-rationalization" techniques to ensure agents follow strict planning, testing, and verification standards.

// ANALYSIS

Agent Skills shifts the focus from "better models" to "better processes," forcing AI agents to act like senior engineers rather than fast-moving juniors.

–Introduces "anti-rationalization" tables that counteract common LLM excuses for skipping tests or documentation
–Standardizes seven key slash commands (/spec, /plan, /build, /test, /review, /code-simplify, /ship) for structured development cycles
–Enforces strict test-driven development (TDD) with a 80/15/5 unit-to-E2E testing ratio requirement
–Compatible across the leading AI coding ecosystem, including Claude Code, Cursor, and GitHub Copilot
–With over 18k GitHub stars, it's becoming the de facto standard for reliable AI-generated codebases

// TAGS

agentai-codingdevtoolopen-sourcetestingcode-reviewaddy-osmaniagent-skills

DISCOVERED

45d ago

2026-04-21

PUBLISHED

45d ago

2026-04-21

RELEVANCE

9/ 10

AUTHOR

AICodeKing

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

LAUNCH21m ago

Augment Cosmos exits preview for teams

Augment Code has launched Cosmos out of public preview to coordinate teams of specialized AI agents across the software development lifecycle. The platform maintains a shared context and organizational memory, enabling agents to learn from team patterns and collaborate on complex engineering tasks.

NEWS28m ago

Nous, H, Prime Intellect join Nemotron Coalition

The Nemotron Coalition has welcomed H Company, Nous Research, and Prime Intellect as new members to accelerate open-weight base model development. The new additions contribute specialized capabilities in agentic AI, post-training, reinforcement learning environments, and distributed training infrastructure.

UPDATE29m ago

TanStack AI has introduced host-side Model Context Protocol (MCP) support via the @tanstack/ai-mcp package, allowing developers to expose any MCP server as type-safe tools for AI agents.

TanStack AI now has first-class, host-side Model Context Protocol (MCP) support. The new `@tanstack/ai-mcp` library treats MCP clients as tool factories that return ordinary server tools, which can be fanned out into chat functions across any LLM adapter or agent framework. It supports both standalone clients and pooled clients with automatic name prefixing to prevent collisions, managed lifecycle operations to avoid connection leaks, and three distinct modes of type safety ranging from zero-config runtime discovery to CLI-generated compile-time TypeScript types.