Enable a Repo¶

"AI enable this repo" is the single entry point. Point your agent at a path and it picks the right mode automatically.

Ask the agent¶

"AI enable /path/to/your-project."

The agent first shows an exec summary + cancel gate (informed consent), then detects any existing AI footprint and chooses one of three modes.

The three modes¶

Fresh (Mode A)Migrate (Mode C)Already-Ours (Mode B)

No prior AI footprint. The agent analyses the repo (language, stack, type), harvests durable facts from your docs, and generates a tailored memory/ layer, bootstrap pointers, skill adapters, and the git-hook + CI triggers (forge-aware: GitHub Actions, GitLab CI, or Azure Pipelines) — plus a DRAFT Vision and its confirmation gate.

You may choose a discovery depth: a standard scan (default) or an /init-depth deep analysis (which writes findings into the neutral memory layer, never a vendor steering file).

The target already uses vendor AI files (Cursor, Aider, Continue, Cline, Roo, Windsurf, Copilot, …). The agent folds steering into memory/, converts chat history into dated session logs, and promotes any vendor skills into the neutral agent-skills/ layer — with originals preserved under legacy/, never deleted. A dry-run is offered.

The repo is already enabled. The agent is idempotent — and if the repo is on an older version, it upgrades in place by reconciling against the current target state (v4.35.0) — one diff-and-apply pass plus the few version-gated semantic steps, not a walk of every release in between. A PRE-APPLY step runs before mechanics when it must preserve content that would otherwise be re-copied; reconcile refuses writes until the protected boundary is safe and explicitly confirmed.

How it converges (v4.35.0)¶

The mechanical ~80% of an enable — verbatim protocol docs, the seven built-in skills, hooks, forge CI, managed .gitignore/.gitattributes blocks — is declared in a tool-side install manifest and applied by a runnable reconcile helper in one pass (dry-run first; the report doubles as the consent summary). The agent's time goes where judgment lives: analyzing your repo, harvesting your docs, seeding memory/, and the DRAFT-Vision gate. On a fresh repo that turns a >10-minute stepwise install into seconds of mechanics plus a few minutes of real analysis.

What lands in the repo¶

Path	Purpose
`memory/`	protocol, continuity, vision, sessions, archive, decay-policy
`AGENTS.md`	the one-line universal shim to `memory/PROTOCOL.md`
`agent-skills/`	seven built-ins + any promoted skills
`.githooks/` + the forge CI config (`.github/workflows/`, `.gitlab-ci.yml` + `.gitlab/`, or `.azuredevops/`)	the ritual triggers (agent-activated, forge-aware)
`.agent/version.md`	version stamp (gates upgrades)
`legacy/`	preserved originals (migration only)

After enabling¶

cd /path/to/your-project
git add . && git commit -m "chore: AI-enable repo"

Then just work — see Getting Started, step 3.

Scope guarantee

Enable only ever creates or modifies files within the target repo root. It never touches ~, ~/.claude/, Application Support, AppData, or system paths.