AI Coding / Vibe Coding

AI Coding Tools, Models, and Workflow Guide

Start from your intent: pick a tool, choose a model, ship AI-built work safely, or learn vibe coding without losing control.

Choose a tool Choose a model Ship AI-built work Vibe coding safely

Last reviewed: 2026-07-17. Tool and model listings show per-item review dates; rankings use dated snapshots.

Start from your intent

Four clear paths into the hub. Each path states who it is for, what it does not replace, and where to go next.

I want to choose an AI coding tool

Pick by workflow: editor, terminal agent, coding agent, or cloud builder—then compare Cursor, Claude Code, and Codex.

Best for: Developers and teams selecting a primary AI coding setup.

Limits: Not a permanent #1 ranking. Product surfaces change; verify official docs.

Tool selection guide Browse tools directory

I want to choose a coding model

Read dated leaderboard snapshots for cost, latency, and context—without treating scores as forever quality ranks.

Best for: Builders comparing models inside an existing tool workflow.

Limits: Public benchmarks are not a private coding exam for your repo.

Model methodology guide Open model rankings

I already used AI to write a project

Verify diffs, builds, secrets, and release readiness with local checklists before public users arrive.

Best for: Anyone shipping AI-generated or AI-assisted changes.

Limits: Checklists do not upload your repo or auto-judge code quality.

Ship checklist Workflow guide

I am a non-technical builder

Break ideas into slices, control permissions and data, accept work against criteria, and plan deploy plus rollback.

Best for: Founders, PMs, and builders using AI app builders or agents.

Limits: Does not replace security review for payments or sensitive data.

Vibe coding guide Project rules overview

AI coding model rankings

See top 50 models

Anthropic - Claude Opus - 5 (max)

Claude Opus 5 max leads this snapshot with a top Intelligence Index of 63 and 1M context.

Click for model details

Context: 1M
AA Index: 63
Cost per task: $2.34
Speed: 52 tok/s

Best for: Frontier repository analysis, complex planning, and high-value agent tasks that need top reasoning.

Anthropic - Claude Opus - 5 (xhigh)

Ties the top Intelligence Index with lower cost per task and much faster first-chunk latency than max.

Click for model details

Context: 1M
AA Index: 63
Cost per task: $1.80
Speed: 52 tok/s

Best for: Deep multi-file edits and coding agents that still need frontier quality with better latency.

Anthropic - Claude Fable - 5

Still among the highest Intelligence Index scores, with a 1M context window and strong output speed.

Click for model details

Context: 1M
AA Index: 62
Cost per task: $3.14
Speed: 63 tok/s

Best for: Deep repository analysis, complex planning, and high-value agent tasks where latency is acceptable.

Anthropic - Claude Opus - 5 (high)

High-reasoning Opus 5 mode with much lower latency than max/xhigh while staying near the top.

Click for model details

Context: 1M
AA Index: 61
Cost per task: $1.23
Speed: 51 tok/s

Best for: Daily Claude agent coding, code review, and tool-heavy workflows that still need strong reasoning.

OpenAI - GPT Sol - 5.6 (max)

Top OpenAI reasoning tier in this snapshot, with 1M context and strong Intelligence Index.

Click for model details

Context: 1M
AA Index: 61
Cost per task: $1.23
Speed: 62 tok/s

Best for: OpenAI-centered coding agents, long-context planning, and max-reasoning software tasks.

SpaceXAI - Grok - 4.6 (high)

New Grok 4.6 high entry near the top, with 500k context and lower cost per task than the frontier Claude/OpenAI max tiers.

Click for model details

Context: 500k
AA Index: 61
Cost per task: $0.84
Speed: 66 tok/s

Best for: xAI-stack coding agents, medium-context planning, and cost-aware frontier experiments.

Recommended AI coding tools

See more AI coding tools

AI coding IDE

Cursor

Cursor combines an editor, chat, agent runs, inline edits, and codebase context for day-to-day software development.

Best for: Developers and teams that want AI assistance inside the editor.

Watch outs: Evaluate rule files, repository indexing, pricing, and how much control the team needs over model behavior.

Read introductionLast reviewed: 2026-07-19 Terminal agent

Claude Code

Claude Code is designed for agentic editing, explanation, debugging, and command-line development loops.

Best for: Engineers comfortable working from a terminal and reviewing diffs.

Watch outs: Check command permissions, repository boundaries, and how the team records agent decisions.

Read introductionLast reviewed: 2026-07-19 Coding agent

Codex

Codex represents OpenAI's coding-agent workflow across app, terminal, editor, and multi-agent development surfaces.

Best for: Teams already building around OpenAI models and agent workflows.

Watch outs: Confirm the exact product surface, model availability, and integration path before adoption.

Read introductionLast reviewed: 2026-07-19 Agentic IDE

Windsurf

Windsurf focuses on AI-assisted coding inside an IDE, with agent flows and codebase-aware development.

Best for: Developers comparing Cursor-style editor agents and team coding environments.

Watch outs: Review product changes, account migration, and roadmap stability when comparing it with other IDEs.

Read introductionLast reviewed: 2026-07-19 AI IDE

Trae

Trae positions itself as an adaptive AI IDE for coding, builder workflows, multimodal input, and project-context assistance.

Best for: Teams evaluating newer AI IDEs and builder-style coding environments.

Watch outs: Check regional availability, platform support, model choices, and data handling before standardizing.

Read introductionLast reviewed: 2026-07-19 Terminal agent

opencode

opencode focuses on terminal-based AI coding with a lightweight interface and model-provider choice.

Best for: Developers comparing newer terminal agents.

Watch outs: Validate provider setup, local permissions, and repository safeguards.

Read introductionLast reviewed: 2026-07-19

Third-party plugins and skills

See more plugins

Reviewed repositories with at least 10,000 GitHub Stars are assigned to one primary capability and ordered by a dated snapshot. Secondary capabilities appear only on the detail page. Stars show attention, not an absolute quality or security score.

#01Coding & implementationCollection

Superpowers

A high-signal Claude Code extension project for skills, workflow upgrades, and agent coding habits.

Claude Code users who want battle-tested workflow extensions.GitHub Stars260.1K2026-07-24Read introductionLast reviewed: 2026-07-19

#02Coding & implementationCollection

ECC

A large engineering system with skills, rules, agents, memory, security, and research-first workflows for major coding agents.

Developers who want a comprehensive, configurable coding-agent setup across languages and tools.GitHub Stars232.6K2026-07-24Read introductionLast reviewed: 2026-07-19

#03Context, rules & integrationsWorkflow

Hermes Agent

Nous Research’s open agent stack that keeps context, tools, and workflows across sessions instead of resetting every chat.

Developers who want a long-running coding agent with integrations, memory, and room to grow custom skills.GitHub Stars219.5K2026-07-24Read introductionLast reviewed: 2026-07-19

#04Planning, requirements & specsSkill

Grill Me

Matt Pocock's viral /grill-me skill: one-question-at-a-time interviews that stress-test a plan before any code is written.

Developers who want to surface hidden assumptions in a design or feature plan before agentic coding starts.GitHub Stars184.6K2026-07-24Read introductionLast reviewed: 2026-07-19

#05Coding & implementationCollection

Anthropic Skills

Anthropic's public skills repository for Claude agents: reusable workflows, document skills, and reference patterns.

Teams standardizing Claude Code, Cowork, or API agents with checked-in skills.GitHub Stars163.7K2026-07-24Read introductionLast reviewed: 2026-07-19

#06Agent orchestration & automationPlugin

gstack

A Claude Code plugin with opinionated product, design, engineering-management, release, documentation, and QA roles.

Founders and small teams that want forceful reviews across the full product-development loop.GitHub Stars124K2026-07-24Read introductionLast reviewed: 2026-07-19

AI coding courses

Published course

AI Literacy for Developers

Practical AI basics for developer workflows: models, prompts, coding assistants, verification, privacy, and reliable AI-assisted work.

7 lessonsOpen course

Planned course series

Planned

AI Coding Tools in Practice

Project setup, rules, permissions, plugins, and review workflows for Cursor, Claude Code, Codex, and OpenCode.

Planned

Vibe Coding Product Practice

From idea and prototype to requirements, boundaries, launch checks, and responsible publishing.

Planned

AI Agent Engineering

Specs, subagents, MCP, tests, PR review, release checks, and reusable team workflows.

AI Coding guides

Browse AI Coding coverage by topic: tool selection, coding models, project setup, workflows, vibe coding, and market updates.