aidlc-doctor

Report how compliant a repo is against the 18-item AI-native template, without changing anything. The read-only companion to aidlc-init.

<Use_When>

• /codepresso:aidlc-doctor <target-path> invoked
• User asks "how AI-native is this repo", "check compliance", "what's missing" </Use_When>

1. Resolve `` (default = current project). 2. If `.codepresso/state/aidlc-profile.json` exists, pass it: `node "${CLAUDE_PLUGIN_ROOT}/scripts/aidlc-cli.mjs" scan --profile ` (honest na-scoping for the team's tools/host). Otherwise run `... scan ` (baseline — everything scored). 3. Show the 18-item table (status + evidence + reason) and the overall %. 4. If `secrets[]` is non-empty → 🔴 surface (masked), recommend rotation. Never echo the raw value. 5. List the top missing/partial gaps, ordered by value. Because scoring is **functional**, explicitly explain items that look present but score lower: - `pre-push` partial → script exists but not wired to a git hook. - `codesight` partial → index exists but is stale and has no regen hook. - `ci-pr` partial → a workflow exists but doesn't match the detected host or has no PR/MR trigger. Mark gaps the scaffolder cannot auto-create (`hooks`, `unit-tests`, `runbook`) as **manual setup**. Suggest `/codepresso:aidlc-init ` to fix. 6. Do NOT write or modify any file.

<Tool_Usage>

• Bash for aidlc-cli.mjs scan (with --profile when a profile exists)
• (read-only — no Write) </Tool_Usage>