intelligence-install-adapter

name	intelligence-install-adapter
description	Enable IDE adapter for intelligence-sync
argument-hint	<target-name>

Install Adapter

The umbrella is whatever directory holds config.yaml (intelligence/, Intelligence/, a codename — never assume the name or casing); the engine module is the directory under it holding scripts/sync.sh (conventionally sync/).

Steps

1. Check whether target $ARGUMENTS is already enabled in config.yaml — if yes, report and stop.

2. Locate the adapter. sync.sh discovers adapters in two places:

   • Built-in: <module>/scripts/adapters/$ARGUMENTS.sh — upstream-owned. update.sh replaces that whole directory on every engine update.
   • Project-owned: <umbrella>/adapters/$ARGUMENTS.sh — update.sh never touches it. A project adapter of the same name overrides the built-in.

   If neither exists, research the tool's prompt format (web search for its rules / agents / skills file layout), then copy <module>/scripts/adapters/_template.sh to <umbrella>/adapters/$ARGUMENTS.sh and implement sync_to_$ARGUMENTS(). Author it there, never inside the engine's scripts/adapters/ — a file written there is deleted by the next engine update. Adapter contract: <module>/docs/ADAPTERS.md. An adapter that would serve every project is worth contributing upstream.

3. Update config.yaml:

   • Target exists with enabled: false → flip to enabled: true.
   • Target missing → add it under targets: with its output: path.
   • If the target reads always-on rules from AGENTS.md (cursor, copilot, codex, pi, opencode), targets.agents must be enabled too — sync fails closed otherwise.

4. Add the adapter's generated paths to .gitignore (the paths it writes, not the whole output root — a shared root like .github/ or .claude/ also holds tracked, hand-authored files).

5. Run /intelligence-sync to generate the output.

6. Report: adapter enabled, files generated, output location.