Skip to content

Choose Plugin or CLI

Archcore has two entry points: the plugin and the CLI. Both read and write the same .archcore/ directory. The plugin wraps the CLI in slash commands and guardrails; the CLI gives you the MCP tools directly.

  • Using Claude Code, Cursor, Codex CLI, or GitHub Copilot CLI? Start with the Plugin. You get skills, guided plan flows, built-in agents, and guardrails. (Codex CLI also has a CLI-only fallback if you’d rather not enable Codex’s hooks feature.)
  • Using any other agent (Gemini CLI, OpenCode, Roo Code, Cline), or scripting in CI? Use the CLI directly.
  • Want to understand the model first? Read the shared concepts.
PluginCLI
Install methodInstall CLI globally, then add the plugin from the marketplaceSingle binary via curl / PowerShell
CLI prerequisiteRequired (same one-line installer as standalone CLI)Required
HostsClaude Code, Cursor, Codex CLI, GitHub Copilot CLIAll 8 supported agents
Interface4 slash commands (/archcore:init, /archcore:plan, /archcore:document, /archcore:review)MCP tools + natural language
Multi-step tracks/archcore:plan tracks and /archcore:document cascades orchestrate document chains automaticallyYou drive them manually through MCP
GuardrailsMCP-only writes, post-mutation validate, cascade warningsMCP available; guardrails are opt-in via hooks
Best forDay-to-day use inside a supported hostScripting, CI, agents without plugin support

Both paths share the same .archcore/ directory and the same MCP server backed by the globally-installed archcore binary. You can switch paths, or use both at once, without migrating anything.

  • CLI is the engine. It reads .archcore/, builds the graph, speaks MCP.
  • Plugin is the runtime on top of it. It applies that graph inside Claude Code, Cursor, Codex CLI, and GitHub Copilot CLI through skills, slash commands, gated tracks, and hooks.

More depth in Mental Model.