Stats
Actions
Tags
From opencode
Internal helper contract for calling OpenCode from the rescue agent.
How this skill is triggered — by the user, by Claude, or both
Slash command
/opencode:opencode-cli-runtimeThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
the ONLY valid command is `task`. there is no `run` subcommand. there is no `--prompt` flag.
the ONLY valid command is task. there is no run subcommand. there is no --prompt flag.
node "${CLAUDE_PLUGIN_ROOT}/scripts/opencode-companion.mjs" task -- "the task text here"
with options:
node "${CLAUDE_PLUGIN_ROOT}/scripts/opencode-companion.mjs" task --model provider/model --agent agent-name -- "the task text here"
# simple task
node "${CLAUDE_PLUGIN_ROOT}/scripts/opencode-companion.mjs" task -- "Say hi to the user"
# with model
node "${CLAUDE_PLUGIN_ROOT}/scripts/opencode-companion.mjs" task --model google/gemini-2.5-pro -- "Explain this code"
# with agent
node "${CLAUDE_PLUGIN_ROOT}/scripts/opencode-companion.mjs" task --agent code-review -- "Review the auth module"
# resume last session
node "${CLAUDE_PLUGIN_ROOT}/scripts/opencode-companion.mjs" task --resume -- "Continue the investigation"
# resume specific session
node "${CLAUDE_PLUGIN_ROOT}/scripts/opencode-companion.mjs" task --session abc123 -- "Pick up from here"
# background
node "${CLAUDE_PLUGIN_ROOT}/scripts/opencode-companion.mjs" task --background -- "Fix the failing tests"
--. everything before -- is flags, everything after is the prompt.task invocation per rescue call.--model is not specified, opencode uses its configured default model.--agent is specified, opencode uses that agent.--resume continues the last opencode session. --session <id> continues a specific one.--fresh forces a new session (or just omit all session flags).--prompt, --quiet, -q, or any other flags not listed above.npx claudepluginhub schmayterling/opencode-plugin-cc --plugin opencodeCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.