/doctor

Orient · Tooling check and hook wiring

Probes declared branch prerequisites, reports known platform install recipes without auto-installing, and wires optional Claude Code hooks.

npx skills@latest add luci-efe/skills -s doctor

OrientStage
06 Autonomy with ownershipMethod
stdout report · .claude/settings.local.json (when wiring hooks)Produced artifact / output

Use when

Setting up, or a skill reports a missing tool, repairing discovery, or configuring the Claude hook adapter.

01
Inspecting branch readiness
Run node <this skill dir>/scripts/doctor.mjs (pass provider IDs for a single branch, or --json for automation; missing/failed probe exits 2)
→ Selected provider matrix on stdout
02
Missing provider binary
Display the verified platform install recipe from references/providers.md; never guess or auto-install
→ Actionable diagnostic, fail fast
03
Wiring Claude Code hooks
Run --install-hooks claude --dry-run first, inspect merge, then register SessionStart, UserPromptSubmit, and Stop in .claude/settings.local.json
→ Unrelated settings preserved; hooks installed (the /proof Stop gate arms only after a passing /proof run in that session)

Selected provider matrix is reported on stdout and missing prerequisites include actionable diagnostics.
Hook installation preserves unrelated settings in .claude/settings.local.json and only Stop blocks claimed proof completion.
Never claims every provider, service, browser, or desktop daemon is operational from a binary probe alone.

Want something like this built for your company?