opencode-build-plugins

Create custom OpenCode plugins using the @opencode-ai/plugin SDK.

9|Updated Feb 21, 2026
One-click install
npx skills add https://github.com/pantheon-org/tekhne --skill opencode-build-plugins
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: opencode-build-plugins
Source: https://github.com/pantheon-org/tekhne/tree/main/skills/agentic-harness/opencode-toolkit/build-plugins
Command: npx skills add https://github.com/pantheon-org/tekhne --skill opencode-build-plugins

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill empowers users to extend OpenCode's functionality by creating custom plugins, enabling advanced automation, custom tool integration, and fine-grained control over AI behavior.

Core Features & Use Cases

  • Plugin Development: Guides users through creating custom plugins using the @opencode-ai/plugin SDK.
  • Hook Implementation: Supports various hooks like tool.execute.before, event, chat.params, and auth for customizing AI actions.
  • Tool Creation: Allows defining new tools that the AI can invoke.
  • Use Case: A user wants to automatically block any rm -rf commands executed by the AI, log all tool calls for auditing, or add a custom tool to look up JIRA tickets.

Quick Start

Create a plugin that blocks dangerous bash commands by throwing an error in the tool.execute.before hook.

Frequently Asked Questions about opencode-build-plugins

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

FAQPage Schema
How do I build custom plugins for the OpenCode AI platform?

To build custom plugins for the OpenCode AI platform, use the `@opencode-ai/plugin` SDK to define custom tools, implement hooks for tool executions, react to lifecycle events, and set up custom authentication flows.

Can I intercept and block dangerous bash commands executed by AI?

Yes, you can intercept dangerous bash commands by implementing the `tool.execute.before` hook in your plugin to check commands and throw an error to block executions like `rm -rf`.

How does hook implementation work for customizing AI actions in OpenCode?

Hook implementation works by attaching custom logic to lifecycle events using supported hooks like `tool.execute.before`, `event`, `chat.params`, and `auth` to intercept executions, log audits, or modify parameters.

Do I need the plugin SDK to add a custom tool that looks up external tickets?

Yes, you need the `@opencode-ai/plugin` SDK to create custom tools that integrate external services, enabling the AI to invoke specific functions like looking up JIRA tickets during a session.

What are the limitations when extending AI capabilities with custom tools?

Limitations when extending AI capabilities with custom tools include relying strictly on the supported hooks provided by the SDK, which may restrict highly complex custom authentication flows or unsupported event reactions.