converting-slash-commands

Converting Claude Code slash commands to OpenCode format involves running the convert_command.sh script on your command files. The script translates Claude Code frontmatter (description, agent, model) to OpenCode equivalents, converts allowed-tools to agent selection, and preserves command content, placeholders, and shell syntax while outputting OpenCode command files to .opencode/command or ~/.config/opencode/command directories.

Command arguments and shell commands are fully preserved during migration. Placeholders like $ARGUMENTS, $1, and $2 remain intact, and shell command syntax is maintained. The conversion focuses on updating the frontmatter structure and field mappings while keeping your command logic unchanged.

Yes, the conversion script supports batch migration across both project-level commands in .claude/commands and user-level commands in ~/.claude/commands. Both are converted to their respective OpenCode locations (.opencode/command and ~/.config/opencode/command) in a single operation.

Fields unsupported by OpenCode are removed during conversion. The script maps supported fields like description, agent, and model to OpenCode conventions, integrates argument hints into descriptions, and handles allowed-tools by applying them as agent selection and permissions rather than transferring them directly.

Yes, after running the convert_command.sh script, review the generated OpenCode command files and test them by invoking the commands. This ensures the frontmatter mapping, field conversions, and argument handling translated correctly for your OpenCode environment.

Claude Code and OpenCode use different frontmatter structures and field naming conventions. This migration Skill translates Claude Code's command definition format to OpenCode's requirements, including how tools and permissions are specified, model selection is declared, and agents are configured.