Plugin Overview
Plugin
The Archcore Plugin makes your AI code like it already knows your repo. It runs on top of the Archcore CLI and brings the architecture, rules, and past decisions of .archcore/ into your agent — automatically, before each edit.
What you get
Section titled “What you get”- 7 outcome-focused slash commands —
/archcore:init,/archcore:context,/archcore:capture,/archcore:plan,/archcore:decide,/archcore:audit,/archcore:help. Each one does one thing and tells you up front what you get from it. - Built-in agents — a universal assistant and a read-only auditor.
- Automatic context — relevant rules and ADRs are injected before the agent edits a source file. No prompting required.
- Guardrails —
.archcore/writes go through MCP only, every mutation is validated, and stale dependents are flagged on cascade.
When to use the plugin
Section titled “When to use the plugin”- You work in Claude Code, Cursor, or Codex CLI.
- You want guided flows and tracks instead of explaining document types to the agent.
- You want context injection and validation to run on their own.
If you’re on a different agent or want a minimal, scriptable install, start with the CLI instead.
Next steps
Section titled “Next steps”- Install the plugin — Claude Code, Cursor, or Codex CLI.
- Supported AI agents — feature parity per host.
- How it works — hooks, injection, MCP-only, validation.
- Skills — all 7 commands at a glance.