agentplugins
Health Warn
- License — License: Apache-2.0
- Description — Repository has a description
- Active repo — Last push 0 days ago
- Low visibility — Only 5 GitHub stars
Code Warn
- process.env — Environment variable access in docs/.vitepress/config.ts
- network request — Outbound network request in docs/package-lock.json
Permissions Pass
- Permissions — No dangerous permissions requested
No AI report is available for this listing yet.
Write AI agent plugins once, ship to multiple agents (Claude Code, OpenCode, Pi, and more). One manifest, any adapters.
AgentPlugins
Write AI agent plugins once, ship to any harness.
Why
- Write once, run everywhere — one manifest compiles to Claude Code, Codex, OpenCode, Pi Mono, Copilot, Gemini, and Kimi
- Universal codegen first, per-harness fallback second — Tier-1 harnesses get full parity; Tier-2 gets skills + commands + a subset of hooks
- No runtime proxy — every adapter emits the harness's native output format; no shim layer, no lock-in
- Honest capability matrix — capability gaps emit
WARNat compile time and surface on the capability matrix page - Plays nicely with what you already have — install a plugin into a harness that already has its own plugins; they coexist
Install
Install the CLI, then add any plugin to every supported AI agent — Tier-1: Claude Code, Codex, OpenCode, Pi Mono. Tier-2: Copilot, Gemini, Kimi.
curl -fsSL https://agentplugins.pages.dev/install.sh | bash
agentplugins add user/awesome-plugin
Or skip the install and use npx:
npx @agentplugins/cli add user/awesome-plugin
Full guide → agentplugins.pages.dev/guide/install
Quick start
From zero to a working plugin in any supported harness:
# Scaffold a new plugin from the official template
npx @agentplugins/cli init my-plugin
cd my-plugin
# Install it into every supported harness (writes to ~/.claude, ~/.codex, etc.)
npx @agentplugins/cli add .
# Verify the install landed where you expect
npx @agentplugins/cli doctor
Create a plugin
Scaffold a plugin from a template, write your manifest, build, and publish to GitHub:
agentplugins init
agentplugins build
Full guide → agentplugins.pages.dev/guide/creating-plugins
Porting an existing plugin? → agentplugins.pages.dev/guide/porting
Supported agents
Tier-1 (full capability parity): Claude Code, Codex, OpenCode, Pi Mono.
Tier-2 (skills + commands, subset of hooks): Copilot, Gemini, Kimi.
Capability comparison → agentplugins.pages.dev/guide/capability-matrix
Architecture
One manifest → seven adapters. Each adapter owns its output format; the @agentplugins/core compiler routes capability expressions to harness-native primitives and emits a WARN for any gap.
Full detail → ARCHITECTURE.md
Documentation
Full docs → agentplugins.pages.dev
LLMs.txt for AI agents → agentplugins.pages.dev/llms.txt
Contributing
PRs welcome. See CONTRIBUTING.md, file issues, and join the conversation in Discussions.
Reviews (0)
Sign in to leave a review.
Leave a reviewNo results found