create-agent-adapter

Create a Paperclip adapter package with server, UI, and CLI entry points.

1|Updated Mar 31, 2026
One-click install
npx skills add https://github.com/kangnam7654/stapler --skill create-agent-adapter-kangnam7654
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: create-agent-adapter
Source: https://github.com/kangnam7654/stapler/tree/main/.agents/skills/create-agent-adapter
Command: npx skills add https://github.com/kangnam7654/stapler --skill create-agent-adapter-kangnam7654

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a technical guide for creating a new Paperclip agent adapter package, detailing the required interfaces, module structure, registration points, and conventions derived from existing adapters like claude-local and codex-local.

Core Features & Use Cases

  • Architectural blueprint for a self-contained adapter package that exposes server, UI, and CLI integrations.
  • Clear guidance on module layout (server, ui, cli, plus index and tests), and how to register with server/ui/cli registries.
  • Instructions for writing a robust agentConfigurationDoc to guide LLM-based agents.

Quick Start

Create a new adapter package under packages/adapters and implement the four entry points, then wire it into server/ui/cli registries.

Frequently Asked Questions about create-agent-adapter

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a new agent adapter package for a TypeScript server and CLI runtime?

To create a new agent adapter package, you build a self-contained module under packages/adapters that implements server execute/parse logic, UI stdout parsing, and CLI formatters following shared adapter-utils contracts. You must export type, label, models, and agentConfigurationDoc from the root index file.

What module structure is required when building a Paperclip adapter?

Building a Paperclip adapter requires structuring your package with dedicated server, ui, and cli modules, plus an index and tests. You must wire each module into its respective server, ui, and cli registries to ensure the runtime, UI config builders, and CLI formatters can consume the integration.

How does an agent adapter integrate with server execute and parse operations?

An agent adapter integrates with server execute and parse operations by implementing the required server-side interfaces defined in the shared adapter-utils contracts. This implementation handles command execution and parses the stdout output to ensure the server runtime can properly manage agent responses.

Do I need to write an agentConfigurationDoc when adding a new CLI integration?

Yes, you need to write an agentConfigurationDoc when adding a new CLI integration. This descriptive documentation is required to guide LLM-based agents and must be exported from the package root alongside the adapter type, label, and models.

Can I use existing adapters like claude-local as a blueprint for a new integration?

Yes, you can use existing adapters like claude-local and codex-local as an architectural blueprint. They provide clear guidance on the required module layout, registration points, and conventions needed to build a self-contained adapter package that exposes server, UI, and CLI integrations.

What are the limitations when building an adapter package without external dependencies?

Building an adapter package without external dependencies requires strictly following the shared adapter-utils contracts for all TypeScript implementations. You must ensure proper registration paths in all server, UI, and CLI registries are maintained manually without external automation libraries.