ZeroSpec is a zero-dependency Markdown framework that helps AI agents understand your repository structure, rules, and source of truth before coding.

SteadyAgent: a bilingual local-first harness for Codex and Claude Code with workflow rules, safety hooks, validation gates, checkpoint commits, and release evidence.

Agent skills for coding CLIs, multi-agent runtimes, context engines, MCP extensions, and terminal tooling. Instead of using claude code's source code, give your agent skills to create your own!

Lightweight fast-path instruction package for AI coding agents: reduces unnecessary context loading and adds state-aware validation.

Workflow for building AI projects with Claude and auditing them with Codex. Claude builds, Codex audits, and Still configures the project.

Tool-agnostic custom instructions, subagent playbooks, skills, and reference docs for production-grade Codex coding agents.

Build curated agent skills for coding CLIs, terminal tools, context engines, remote bridges, and multi-agent runtimes

Multi-Agent Long-Task Scheduling and Growth System for AI coding agents, with file-based state, verification, handoff, and reusable learning.

For people who keep Agent Skills on their machine or team drive. Switch repos without dragging the same pack everywhere: keep one local catalog, then use kit to pick and install the right skills per project so your coding AI gets what that repo needs, without a cloud marketplace.

OpenCode skill that creates, adapts, and audits agent skills from vague user requests. Generates SKILL.md files with spec, audit, and safety boundaries.

Open-source starter kit for agentic AI development. Includes Docker-based dev environment, MCP servers, local and cloud LLM support, VS Code Dev Containers, automation scripts, and AI coding workflows.

Persistent local memory across LLM agents, CLIs, coding tools, and developer workflows.

A compact operating system for coding agents that routes engineering tasks to focused playbooks, minimizes context usage, and enforces evidence-driven implementation and verification.

Shared-memory operating environment for AI coding agents. The repo is the memory: agents wake severed from private context and start from the same record.

Windows/PowerShell safety harness for Codex CLI and Claude Code CLI with dry-run defaults, prompt artifacts, bounded fix loops, and human-reviewed handoffs.

WezTerm config for AI coding agents (claude coder + codex reviewer): keybindings, left-right split, one-click install, and session crash-recovery. Windows.

Bare-repo Git worktree patterns, scripts, and agent guidance for parallel coding workflows

Musters multiple AI coding agents into orderly, evidence-backed work behind deterministic policy gates.

Modern local toolchain setup scripts for coding agents

PASS/STOP pre-action receipts for Codex and AI coding agents before edits, handoffs, commits, or publish steps.