工具与命令

@dietrichgebert/ponytailLazy senior dev mode for AI agents. The best code is the code you never wrote.

@plannotator/opencodePlannotator plugin for OpenCode - interactive plan review with visual annotation

@prevalentware/opencode-goal-pluginOpenCode goal plugin that adds Codex-style long-running goal mode, /goal commands, persistence, and TUI status for AI coding agents.

opencode-ptyOpenCode plugin for interactive PTY management - run background processes, send input, read output with regex filtering

opencode-cmd-providerCommand Code provider + plugin for opencode

opencode-skills-collectionOpenCode CLI plugin that automatically downloads and keeps skills up to date.

opencode-skill-creatorOpenCode plugin for skill creation — custom tools for validation, evaluation, description optimization, benchmarking, and review serving.

jev-guardPrompt-injection and dangerous-action guard for coding agents (Claude Code, Codex, pi, ACP), powered by Jev.

opencode-snipOpenCode plugin that prefixes shell commands with snip to reduce LLM token consumption

@knownasred/plannotator-opencodePlannotator plugin for OpenCode - interactive plan review with visual annotation

opencode-landstripOS-level AI command sandboxing for OpenCode with Landstrip

opencode-skills-tuiOpenCode TUI plugin that shows skills and loaded state in the sidebar

opencode-auto-permissionsAutomatic, context-aware permission review for OpenCode

@imransyafi/opencode-bgOpenCode plugin for interactive PTY management - run background processes, send input, read output with regex filtering

@brainervirus/opencode-commandcodeCommand Code API provider for opencode — use Claude, GPT, Gemini, DeepSeek, Qwen, Kimi, GLM, MiniMax, and Step models via Command Code

korean-report-skills한국어 기술·사업 문서의 자립형 HTML 제작과 문체·프레이밍·정확성 검토를 제공하는 Agent Skills

opencode-queueQueue OpenCode prompts and slash commands until the agent is idle

@wallbreakerno4/opencode-commandcodeOpenCode provider plugin for the Command Code Go plan: an in-process bridge to the /alpha/generate gateway.

opencode-datarobot-skillsOpenCode plugin providing DataRobot skills for AI/ML workflows and a DataRobot-branded theme.

dejavu-gatesCross-session error gates for AI coding agents: detects recurring tool-call failures and promotes them into enforced gates. Remind first, block on same-session repeat offense. Works with OpenCode, Claude Code, Codex CLI, Gemini CLI, Cursor, Copilot CLI, a

opencode-atuin-historyOpenCode plugin that records agent bash commands in Atuin shell history

@mynameistito/opencode-usage-limitsOpenCode TUI plugin that shows Codex, Command Code, OpenCode GO, ZAI, Synthetic, MiniMax Token Plan, and Qwen usage limits in the sidebar and prompt footer.

@edo1994/opencode-queueQueue OpenCode prompts and slash commands until the agent is idle

opencode-sidebar-token-metricsOpenCode sidebar plugin showing generation speed (tok/s), cache hit rate, and average time-to-first-token (TTFT).

opencode-effect-enforcerOpenCode V2 plugin for Effect v4 skills, guidance, and pattern enforcement

opencode-sidebartmux sidebar launcher for OpenCode

opencode-tokenwatchReal-time token usage, cache analytics & performance dashboard plugin for OpenCode CLI

@opencode-cockpit/shellBackground shells for OpenCode: the agent starts, waits on and drives PTYs; you watch them in a docked TUI panel

@himanshujangir/frankLocal receipt and pushback guardrails for AI coding agents.

@warp-dot-dev/opencode-warpWarp terminal integration for OpenCode — native notifications and more

opencode-plugin-loop/loop command for opencode — run prompts on a schedule (fixed, adaptive, or maintenance), modeled after Claude Code's /loop

opencode-thinking-fixRestores reasoning content dropped from multi-turn OpenCode conversations (DeepSeek, Kimi, GLM, MiMo, MiniMax, OpenCode Go, OpenRouter). Plugin + proxy + watchdog.

@prevalentware/opencode-loop-pluginOpenCode loop plugin that adds Claude Code-style /loop recurring prompts: a persistent scheduler that re-injects an instruction into a session on an interval while it is idle.

opencode-plugin-browserSidebar section listing OpenCode plugins with installed/uninstalled status, below MCPs and Skills.

@opencode-cockpit/statusA statusline for OpenCode you can actually configure: declarative segments, a typed module, or your existing Claude Code statusline command

commandcode-go-opencode-providerCommand Code API provider for opencode — use Claude, GPT, Gemini, DeepSeek, Qwen, Kimi, GLM, MiniMax, and Step models via Command Code

@rajnandan1/kenThompson-mode systems discipline for AI agents. Try it, and if it doesn't work, throw it out and do it again.

opencode-tmux-indicatorOpenCode CLI plugin that sets a tmux window option when the agent is waiting for input

opencode-plugin-gstackGStack - Virtual software engineering team with specialized AI skills for OpenCode

@telsoncc/opencode-commandcodeCommand Code API provider for opencode — use Claude, GPT, Gemini, DeepSeek, Qwen, Kimi, GLM, MiniMax, and Step models via Command Code

@yuanbopang/frugalCloud cost awareness plugin for coding agents (Claude Code, Codex, Copilot, Qoder, Gemini CLI, OpenCode, pi, Cursor, Windsurf, Cline, Kiro)

opencode-terminal-progressOpenCode CLI plugin that shows agent progress in terminal tabs (Ghostty, iTerm2, WezTerm, Windows Terminal)

@opencode-cockpit/updaterEvery OpenCode plugin you have installed: what actually runs, what is published, and an update that is checked against disk

@ursuciprian/reflexProd-safe AI coding agents for infra teams: terraform AI agent guardrails (plan-aware apply gate), AWS, kubectl and Terraform workspace context, change freezes, a team policy and a SOC 2 / ISO 27001 audit export for Claude Code, Codex CLI, opencode and pi

@spoons-and-mirrors/subtask2Opencode plugin for enhanced subtask control with return context and prompt chaining

@openspoon/subtask2Opencode plugin for enhanced subtask control with return context and prompt chaining

opencode-seatbeltmacOS Seatbelt (kernel sandbox) for OpenCode. Stops coding agents from reading secrets, enforced by the OS, not by command matching.

opencode-command-injectAuto-inject project commands into OpenCode. Finds Makefile targets, package.json scripts, and discovered local skills at startup.

@justsilver/opencode-goal-pluginCodex/OMP-style persistent goals for OpenCode V2: /goal command + goal tool + idle auto-continuation + evidence-based completion + blocked/budget guards

@geohar/opencode-sharedserverOpenCode plugin: manage shared backend processes via the sharedserver CLI.

opencode-ssh-sessionOpenCode plugin for persistent SSH sessions with shell state persistence

@herouucn/opencode-commandcodeInject Command Code API provider into opencode — access Claude, GPT, Gemini, DeepSeek, Qwen, Kimi, GLM, MiniMax & Step via unified Command Code API. / 将 Command Code API provider 注入 opencode —— 通过统一的 Command Code API 使用 Claude、GPT、Gemini、DeepSeek、Qwen、Kim

opencode-tsiftOpenCode command shortcuts for tsift bounded code navigation and session evidence

no-vibeTutor mode for AI coding assistants. AI shows code in chat and reviews your work, but never writes to your project files. Ships across Claude Code, OpenCode, Codex, Gemini CLI, and Pi.

opencode-usage-reportopencode plugin: /usage command showing subscription quota windows for configured providers

@brutalsystems/birddog-opencodeopencode plugin for birddog — publishes a session's state so birddog can observe it from outside. Read-only: it reports, and never acts on the session.

opencode-triageDeterministic skill router for OpenCode — saves ~90% tokens by hiding skills from the system prompt and routing via keyword matching

planweftPersistent planning, project documentation and evidence for coding agents

gutchapa-opencode-telegramTelegram bot for opencode: chat with your local AI assistant over Telegram, with 70+ slash commands

opencode-rundevRun & Debug for the terminal-first world: bring a repo's dev environment up and down from OpenCode — containers, dev servers, emulators — with no LLM in the loop.