沙箱与基础设施

opencode-sandboxOpenCode plugin that sandboxes agent commands using @anthropic-ai/sandbox-runtime (seatbelt on macOS, bubblewrap on Linux)

@skyvence/coolify-ocDeploy and configure projects on a self-hosted Coolify instance from OpenCode: model-facing tools plus a separately installed sidebar.

@konfeature/opencode-atelierOpenCode plugin that dispatches coding tasks to Atelier sandboxes (self-hosted Kata Containers dev environments).

opencode-shipit-runtimeOpenCode plugin exposing Ship-It build, runtime, artifact, filesystem and usage tools to the AI

opencode-vision-analyzeOpenCode plugin: vision_analyze tool — describe images & image URLs with your own vision model for text-only main models; native-image fast path for multimodal models. Zero runtime dependencies.

oc-mediaOpenCode plugin: teaches the native Read tool to ingest audio and video files, gated live by each model's real input modalities from models.dev.

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

@veris-ai/daytona-opencodeRun every OpenCode session in a Daytona sandbox whose vendor API calls are answered by a Veris twin. One line in opencode.json: unmodified code, real hostnames, and a receipt of what the vendor actually received.

@veris-ai/e2b-opencodeOpenCode plugin that runs every session in an E2B sandbox whose vendor API calls are answered by a Veris twin, with a receipt of what the vendor received.

opencode-analyze-imageOpenCode plugin that lets text-only models analyze image attachments, URLs, and files through an auxiliary vision model.

@we64/opencode-vision-multimodalGive opencode native image, PDF and document analysis via any OpenAI-compatible multimodal API. Files are sent as-is; the API decides what it can read. No OCR, no conversion.

@jsonzx/opencode-multimodal-attachmentsGive text-only OpenCode models multimodal attachment analysis through Gemini.

opencode-vision-fallbackOpenCode plugin: auto-describe images via a vision model when the active model lacks vision support.

opencode-eyeRoute image inspection to an eye-capable model when the active model cannot see images

opencode-media-studioNative video generation, audio, and video inputs for OpenCode

opencode-sandbox-winOpenCode plugin that sandboxes agent commands using @anthropic-ai/sandbox-runtime (seatbelt on macOS, bubblewrap on Linux)

opencode-k8s-workspacesRun remote OpenCode workspaces in Kubernetes

opencode-openchamber-mobile-bridgeOpenCode helper package for exposing OpenChamber from devcontainers through a host bridge and Tailscale Serve.

@opencoins/launchpad-pluginToken launchpad service for OpenCode.ai - Deploy EVM and Solana tokens with AI assistance