OPCStudio — Development skill for Claude Code
Local-first multi-agent company and project work system with evidence-backed delivery, governed memory, MCP, ACP, Codex and Claude integrations.
How to install OPCStudio
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open WUBING2023/OPCStudio and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What OPCStudio does
Local-first multi-agent company and project work system with evidence-backed delivery, governed memory, MCP, ACP, Codex and Claude integrations.
Alternatives in Development
- IPolloWork — Enterprise-grade, local-first Agent Workbench for people and agent teams 4.9k ★
- Lacp — Local-first policy, evidence, memory, and recovery controls for Claude, Codex, Hermes, and other CLI coding ag 300 ★
- Acp Go SDK — Go SDK for the Agent Client Protocol (ACP), offering typed requests, responses, and helpers so Go applications 226 ★
README
OPC Studio
**A local-first, verifiable work system for long-running AI teams.**
[English](./README.md) · [简体中文](./README_ZH.md) · [日本語](./README_JA.md) · [Deutsch](./README_DE.md)
[](https://github.com/WUBING2023/OPCStudio/stargazers) [](https://github.com/WUBING2023/OPCStudio/releases/latest) [](https://github.com/WUBING2023/OPCStudio/releases/latest) [](./LICENSE)
[Website](https://opcstudio.pages.dev/) · [Download](https://github.com/WUBING2023/OPCStudio/releases/latest) · [Documentation](#documentation) · [Report an issue](https://github.com/WUBING2023/OPCStudio/issues)

What is OPC Studio?
OPC Studio organizes API models and subscription CLIs into reusable AI companies. A company defines long-lived roles, responsibilities, permissions, tools, verification rules, and governed memory. Each mission creates a right-sized task graph and a temporary execution team.
The goal is not to make more agents talk. The goal is to make AI work **traceable, verifiable, reusable, and honest about failure**.
**Release channel:** Windows Private Alpha. Use non-sensitive test projects until you have reviewed provider permissions and the [security boundary](./docs/security-boundary.md).
Core model
Organization layer Roles · responsibilities · permissions · Skills · MCP · memory
Task layer Mission graph · dependencies · delivery contract · approval points
Execution layer Model sessions · worktrees · tools · temporary A2A messages
Evidence layer Artifacts · hashes
Related Skills
Do SOUL Alaya
The local-first memory plane for CLI coding agents. Codex and Claude Code attach via MCP; SQLite-backed, evide
Claude Code Cockpit
Local-first VSCode HUD for Claude Code — multi-period token+cost rollups, cross-project session watchtower, Ma
Recallnest
One memory, every terminal. Shared memory layer for any MCP client — Claude Code, Codex, Kimi, Antigravity (ag
Eigen
Opinionated, terminal-first local coding agent for Go/Linux: one daemon, many resumable sessions, multi-provid
Bullpen Init
Silent first-run setup. No questions. Initializes the local memory store, writes sensible defaults, prints a o
Vault OS
Local-first Obsidian memory OS for Claude Code and Codex: trust-aware L0-L3 retrieval, isolated zones, 150 ski
Related Agents
Toolchain Installer
Use for requests to install, configure, repair, or validate MCP servers, Claude Code skills, tool integrations
Mb Researcher
Research specialist for Memory Bank workflows. Use for ecosystem research, implementation reconnaissance, sour
Start Here
Use this agent on first run to set up a new workspace from scratch. Walks the user through identity, agent sel