AIAny
AI Agent2026
Icon for item

Flutter Agent Skills

Provides a set of task-focused agent “skills” — small folders of instructions that teach agents how to perform common Flutter development workflows (integration tests, widget previews, routing, localization). Maintained by the Flutter team to reduce mistakes and make repeatable dev tasks reliable.

Introduction

As language agents are asked to help with app development, the gap is rarely tooling access — it's domain knowledge about ‘how’ to perform repeatable developer workflows. This repository encodes common Flutter developer workflows as small, opinionated "skills" (folders of instructions) so agents can follow reliable, happy-path procedures when adding tests, previews, routing, localization, and other tasks.

What Sets It Apart
  • Skills-as-instructions: each skill is a self-contained folder describing a concrete developer workflow (e.g., add an integration test, create widget previews). So what: agents get step-by-step domain expertise rather than guessing how to use tools.
  • Workflow-focused rather than API-focused: the repo prioritizes repeatable developer flows (testing, layout fixes, serialization, routing). So what: it’s optimized for practical tasks maintainers perform frequently, not for exposing low-level APIs.
  • Maintained by the Flutter team and designed to plug into agent frameworks (the README references MCP and an agent-install UX). So what: it’s opinionated and aligned with Flutter best practices, lowering the chance of agents producing unsafe or non-idiomatic edits.
  • Lightweight, composable units: skills are small and focused, enabling selective installation and easier updates. So what: teams can adopt only the workflows they need without large tooling changes.
Who It's For & Tradeoffs

Great fit if your project uses agents to automate or assist Flutter development tasks, you already run an agent framework that supports skill installation (for example MCP-like systems), and you value predictable, repeatable edits over exploratory code changes. Look elsewhere if you expect the agent to perform deep, project-specific refactors or business-logic design without human review — skills target the happy path and explicit workflows, not open-ended architecture decisions. Also note the repo is opinionated (Flutter team conventions) and currently not structured for heavy community PRs (the README indicates they aren't accepting pull requests), so expect to adapt or fork for large custom needs.

Where It Fits

This collection complements agent tooling that provides tool access (MCP-style toolsets) by teaching the agent the “how” for Flutter tasks. Use it to standardize agent-driven dev flows (tests, previews, localization) while keeping final review gates with humans.

Information

  • Websitegithub.com
  • AuthorsFlutter team (Google)
  • Published date2026/02/25

Categories

More Items

Hugging Face
AI Model2026

Open-weights LLM fine-tuned for phone-based voice agents that prioritizes low latency and reliable tool/function calling. Based on NVIDIA Nemotron 3 Nano (30B total, 3.5B active), supports very long contexts (262,144 tokens) and recommends temperature=0 with thinking disabled for deployment.

GitHub
AI Agent2026

A curated collection of production-ready Agent Skills that turn tasks—presentation production, image generation, local KB retrieval, article assembly, and web-design—into agent-loadable skill folders. Uses a SKILL.md contract, supports multiple agent runtimes (Claude Code, Cursor, Codex), and offers modular install paths with pinned release zips.

GitHub
AI Agent2026

Provides a ruleset and skills that make AI coding agents prefer the simplest correct implementation: reuse existing code, prefer stdlib/native features, and only write minimal new code. Cuts generated LOC, tokens, cost and time while preserving validation and safety.