spk-start-here

What it does

Orients you by where you're working — command line or a supported agent harness — then routes to the smallest useful workflow. It names Spec Kitty's three visible layers (commands, skills, doctrine) and points at the right entry-point skill for your situation instead of walking you through a full tutorial.

When to reach for it

Use it as your first stop in a new session: new project, broken install, starting a mission from scratch, unsure whether a command exists as a slash command or a skill, or unsure which spk-* skill applies to what you're trying to do.

Trigger phrase

There is no single fixed invocation string. Harnesses route natural-language intent — "where do I start," "how do I use Spec Kitty," "what should I run first" — to this skill, or you invoke it directly by name (spk-start-here) using your harness's own skill syntax. It is distinct from CLI use: spk-* skills are agent-harness operating guides, not spec-kitty CLI commands.

First-route table

  • New project or broken install → spk-admin-setup-doctor
  • New mission from scratch → spk-start-first-feature
  • Want a command list → spk-start-command-map
  • "Will this work in Codex or Claude?" → spk-start-agent-surface
  • Existing mission needs advancement → spk-run-next
  • Multi-mission or multi-repo program → spk-run-program-orchestrate
  • Review or approval work → spk-run-review-wp, then spk-gate-accept
  • Team, SaaS, tracker, or sync concern → spk-team-sync or spk-team-tracker
  • Doctrine or governance concern → the spk-doctrine-* family
  • Unsure which skill applies → spk-meta-skill-map

What it does not do

It does not turn into a full tutorial and does not do the routed work itself — it states the route, loads the one skill that matches, and stops.

Edit this page