Coding Agents

A natural-language diagram engine with domain-aware templates, style references, and validation checks for AI and agent workflows.

A deterministic Canvas-based system for turning prompts, clip specs, and style recipes into art-directed motion graphics with repeatable timing, editable scenes, and explainer-ready grammar.

How MiniMax-AI stripped the agentic workflow down to its high-reasoning core using interleaved thinking and the Model Context Protocol.

How a bleeding-edge terminal application uses native cookie syncing and AI agents to replace brittle UI tests with declarative intent.

Moving beyond the tabbed IDE to an infinite canvas where AI agents and persistent terminals live as first-class citizens.

Instead of asking an LLM to write a giant blob of HTML, this project has it write a tight draft and lets a local CLI assemble the diagrams, styling, linting, and interactivity.

A local-first workflow where your AI coding agent writes the markup, preserves design through diff edits, and exports the same content to magazine pages, decks, posters, and social formats.

How a containerized workstation solves the "Docker-Chromium" friction and gives Claude Code a permanent home.

Inside the Codex plugin that stores visual state locally, reads red annotations as intent, and uses MCP to let an agent place and revise images with spatial precision.

Leonxlnx reconstructs the layered prompt architecture behind Claude Code, and shows why the real breakthrough is separation, not just scale.

Zero is not trying to be the most pleasant language for humans. It is trying to be the most legible system for LLMs, with structured compiler output, embedded skills, and a syntax surface designed to reduce ambiguity.

A provider-neutral skill for Claude Code and Codex that turns agent design into a runtime problem, not a prompt-writing exercise.

A DeepSeek Harness extension that gives agents a file tree, Git control, task surfaces, and hard security boundaries, without turning the UI into a brittle monolith.

HKUDS's open-source system turns repositories into layered structure, then retrieves just enough context to answer a question without flooding the model.

How a collection of terminal scripts hijacked Claude Code to build a cognitive workbench with forced artistic chaos and Emacs Org-mode.

A tiny Bash, PowerShell, and Node.js bridge that keeps Claude Code’s agent loop intact while rerouting the model layer to cheaper Anthropic-compatible backends.

A deep dive into Ryan Lopopolo’s agent-first system for routing evidence, preserving provenance, and improving AI output by engineering the harness instead of the model.

Official Expo skills turn docs into executable guidance, so AI assistants stop freelancing and start following the same workflows the core team trusts.

How an Anthropic Claude Code skill is turning "black box" repositories into interactive, single-page masterclasses for the next generation of builders.

A semantic CLI for AI agents that filters noisy repositories into high-value code paths, symbol leads, and bounded context before the model burns through its budget.

A deep dive into the local-first agent that narrows tools, shrinks context, compiles prompt logic, and escalates only when the model actually needs help.

An async-first harness for durable agent work, forkable history, and crash-proof tool execution.

How an opinionated Next.js stack uses embedded rules, Server Actions, and strict conventions to turn Cursor and Claude into autonomous product engineers.

A hook-driven, SQLite-backed system for AI coding agents that captures corrections, preserves context, and turns repeated mistakes into reusable rules.

A Claude Code skill that reads your app, extracts what it should sell, and uses a deterministic Python scaffold plus AI polish to ship store-ready assets.

A deep dive into how `cbrock84/headcount` uses departments, reviewer-class agents, and write-surface enforcement to turn Claude Code into a governable system, not a prompt pile.

How Google uses the Model Context Protocol to turn generative UI into a pluggable tool for AI agents.

A zero-build static-HTML slide system that pairs CSS token themes, multi-window presenter mode, and an AI skill manifest into one unusually complete deck-making stack.

A local-first Go app that edits hidden config files surgically, proxies API calls, and lets different coding agents share models without manual hacking.

A spec-driven workflow for orchestrating AI agents with phase gates, atomic commits, and fresh context windows.

From PDF parsing to GitHub enrichment to rubric-based scoring, this project shows how to build an AI evaluator that explains its own verdict.

A reverse-engineered playbook for drawing the line between web UI, native shell, IPC, and Rust so desktop apps stay fast, polished, and hard to fake.

A modular AI skill that distills creator heuristics, algorithm weights, and account diagnostics into a layered workflow that LLMs can actually use.

A framework-less Python demo that turns web search into a report-writing loop: structure first, search second, reflection third, then another search if the draft still has gaps.

It checks prompts, rules, skills, hooks, subagents, and verifiers from inside the agent itself, then tells you what your setup is missing.

A TypeScript generator, a curriculum file, and a single flattened README turn a pile of Markdown terms into a controlled vocabulary for the age of agents, MCP, and vibe coding.

By marrying Claude 3.7 with ephemeral sandboxes and durable workflows, Vercel-Labs is moving beyond LLM "guessing" toward verified software engineering.

A lightweight skill framework that turns a handful of files into a real-time avatar, voice, and context pipeline for autonomous agents.

A terminal-first skill suite that routes audits, copy, emails, and reports through specialist subagents, then turns the results into structured deliverables.

How a clever network proxy and on-the-fly bundle patching turned Anthropic's locked-down terminal client into a universal, local-first AI agent.

A vanishingly small JavaScript engine, a seam-aware generation pipeline, and an AI-agent workflow that makes premium cinematic landing pages feel programmable.

A meta-framework that teaches coding agents to harvest golden paths, reject lucky guesses, and write only the lessons worth keeping.

A Codex Skill that separates emotion, evidence, and action. It rejects mind-reading, routes around danger, and turns messy human conflict into a disciplined decision pipeline.

A local-first autonomous agent that bootstraps its own runtime, browser tooling, and multi-model council from one entry point.

A deep look at the open-source kit that turns parametric modeling into a language agents can actually learn, test, and execute.

A deep dive into how CAR uses tickets, durable context, and resumable flows to make agents behave less like chatbots and more like managed processes.

How a local MCP bridge uses React Fiber traversal and Chrome DevTools to turn a closed-source trading terminal into an AI control plane.

A local-first development environment for autonomous coding, where tasks persist, execution is sandboxed, and every action is designed to be audited.