Browse agent skills

Jev-powered Auto Mode, selectable model router, agent routing, and context filtering for OpenCode 1.18.x.

Real-time OpenCode context-window usage indicator: per-category token breakdown written to a log file, plus an optional TUI sidebar. Dual V1 + V2 plugin entry.

A live status line for OpenCode v2's terminal UI — context window, cache, streaming speed, cost, elapsed time and uncommitted changes in one configurable row.

Sebas, the electronic butler for LLM-assisted work: speaks your agents' messages through a local text-to-speech daemon, parks the messages that collide onto a system notification card with a single Listen button, and keeps one voice at a time. Ships the v

Deterministic multi-agent workflow orchestration and an ultracode effort mode for opencode

OpenCode V2 plugin: deliver audio and video to media-capable models via @-attachments and the read tool

Keyless OpenAI-compatible API (Chat Completions + Responses) for OpenCode's free Zen models, served through the real OpenCode CLI.

Verbatim context pruning for OpenCode v1 and v2: prune stale tool calls and truncate bulky tool results from the outgoing model request using TypeSafe Jev decisions, instead of summarizing context.

TurnPrune for OpenCode: replaces tool outputs the agent no longer needs with a fixed placeholder, once per user turn.

Deterministic multi-agent Workflows for OpenCode: TypeScript scripts that fan work out to subagents, with typed results, a live TUI panel, a web app and a versioned protocol.

OpenCode v2 CLI with prompt modes, code-review-graph auto-indexing, and the browser skill

JEV Orchestrator plugin for OpenCode — O(1) intra-loop tool routing via Typesafe JEV on OpenRouter

Opinionated modular memory system for coding agents (OpenCode + Pi)

Speech-to-text and text-to-speech for OpenCode. Record voice prompts with whisper-cpp, hear responses via Piper TTS, with LLM normalization through any OpenAI-compatible endpoint.

OpenCode V2 plugin providing real-time Token Usage and context window analytics inspired by Kilo Code / Kilo CLI

OpenCode Go provider for DeepSeek Harness: live gateway catalog, per-conversation session header, usage

Let OpenCode operate your desktop apps and Chrome tabs with the Codex Computer Use engine already on your machine

Visual bridge for opencode. Registers the native vision_analyze tool (model configured by the user via the opencode agent model override on vision-agent) with the vision-agent subagent as fallback, and teaches a text-only orchestrator to extract visual in

Typed decisions for the Opencode harness via Jev (System One) or OpenJev equivalent — Choice/Noul/Score with calibrated probabilities, no text generation

Multi-agent workflow toolkit for OpenCode: 10 role-tuned agents, the workflow-driver and workflow-subagent skills, and permission composition that keeps your overrides winning

Chat with local AI coding agents from Telegram, powered by the OpenCode API

Adaptive reasoning effort for OpenCode V2 with request-local GPT-6 model selection

Run one engine attempt through a fresh OpenCode CLI process, with your own OpenCode login.

OpenCode adapter for Seqlane.

OpenCode TUI plugin that renders session context-window usage as a segmented bar in the sidebar

Live inference telemetry pinned to the OpenCode 2 sidebar.

OpenCode V2 image generation PoC using standard provider Images APIs

Free OpenCode Zen models in DeepSeek Harness through native HTTP requests. No OpenCode installation, login, API key, server, container, or LiteLLM is required.

The Windows x64 runtime of the omnirush CLI (bun, fd, ripgrep, opencode). Installed automatically by omnirush.

TypeSafe Jev skill-suggestion plugin for OpenCode: cheap, calibrated skill routing before the main LLM runs