OpenCode routines: same-session loops, cron prompts, and host-backed standalone scheduled agents
Directory
Browse agent skills
Search by intent: design systems, testing, deployment, security, documentation, Azure, Supabase, React, or any workflow you want an agent to perform better.
Showing 1–30 of 602 skills
Page 1 of 21 · loading 30 at a time
Cerebro/X-Men workflow for OpenCode: agent planning + deterministic TypeScript execution engine + Cyclops audit.
OpenCode add-on providing structured multi-agent workflows, durable task artifacts, and an orchestrator agent for complex AI coding sessions.
Hermes-style Telegram bot integration for opencode — forward messages, handle media, approve tool calls via inline buttons
Distributed Cognitive Architecture for OpenCode. Turns simple prompts into specialized multi-agent workflows (Planner, Coder, Reviewer).
Official Haiilo AI plugin for OpenCode — auth (Manual API Key + Okta Device Flow) and automatic model discovery
Swarm DAO governance adapter for OpenCode — multi-agent deliberation, quality gates, and structured decision-making
Lume — Nebular's coding coach for the terminal. A teaching protocol on top of OpenCode that walks learners through projects against a Nebular knowledge graph.
OpenCode TUI sidebar panel showing per-session timing: total API/inference time, wall-clock, turns, average and slowest turn. Zero context-window pollution.
OpenCode TUI plugin displaying DeepSeek account balance in the sidebar
OpenCode plugin for AWS Bedrock token authentication via the OpenAI-compatible endpoint
Voice-to-text plugin for opencode — toggle recording, transcribe with OpenAI Whisper, inject into TUI
Independent multi-agent team orchestration plugin for OpenCode
OpenCode plugin that shows ChatGPT/OpenAI usage in the sidebar and command palette.
LangChain BaseChatModel adapter for @earendil-works/pi-ai — use Pi with any provider, model and auth. Includes a native opencode/Zen chat model (no pi).
OpenCode plugin for local-first RAG-based semantic code search
Prevent, detect, and reduce AI hallucinations in LLM outputs. A comprehensive skill with prompts, techniques, and validation tools.
Telegram bot client for OpenCode to run and monitor coding tasks from chat.
Task persistence, listing, and pipeline chaining for opencode background subagents.
OpenCode plugin that profiles tool calls — tracks duration, frequency, errors, and visualizes stats with Chart.js
OpenCode plugin: repairs open-model tool-call wire-format violations (DeepSeek/Qwen/GLM/Kimi/etc.) via an in-process OpenAI-compatible repair proxy, before opencode's AI SDK validates them.
Every project begins here. Genesis reads your codebase and builds a tailored AI engineering team.
OpenCode plugin for document-to-Markdown conversion via Microsoft MarkItDown
Persistent memory & knowledge graph plugin for OpenCode — remember facts, decisions, preferences across sessions
Multi-agent parallel code/spec review for OpenCode
Vision plugin + tool for OpenCode — automatically handles pasted images for non-vision models
OpenCode imperial multi-agent orchestration plugin.
Multi-agent SDLC harness with knowledge graph, quality gates, and swappable LLM backends