Skip to content

OpenCode & Agents

The repo ships a shared OpenCode configuration: agents, slash commands, skills, and lifecycle plugins. The source lives in agents/.config/opencode/ (agents, commands, plugins) and agents/.agents/skills/ (skills shared by OpenCode and Codex via ~/.agents/skills/).

These are generated from this dotfiles repo and published automatically to timmo001/opencode-config so they can be browsed, imported, and installed independently of the full dotfiles repo.

  • Graduated agent permissions — agents range from fully read-only (reviewer, ask) through ask-gated (build-ask) to edit-capable (refactorer, build-locked). A guard plugin stops read-only agents from escalating via subagent delegation.
  • Skill-based routing — commands are thin wrappers that name an agent, declare required skills, and state whether branch context is needed. The workflow logic lives in skills and plugins.
  • Branch context injection — the branch-context plugin pre-computes git/PR state once per command and injects it as structured XML, so commands do not run their own git/gh calls.
  • Secret protection — the env-protection plugin blocks reads of .env files (except .env.example).
  • dot agents-sync — mirror ~/.config/opencode/AGENTS.md into a Cursor rule.
  • dot opencode-debug [--agent <name>] — inspect resolved OpenCode config, paths, and skills.
  • dot skill-updates / dot skill-check — check/apply upstream skill updates and validate skill references.

Publishing to opencode-config happens automatically via GitHub Actions on push. The Agents, Commands, Skills, and Plugins pages above are generated by bun run gen:opencode from the live assets, so they always match what ships — edit the source assets, not the generated pages. See also the generated dot Command Reference.