Create a verification skill

Every serious project needs a scripted way to drive the real app and prove behavior: launch it, exercise a feature the way a user would, and capture evidence. This skill generates that as a project-local skill (.omp/skills/verify-<app>/) tailored to the repo. You write the generator's output for the next agent, not for a human: it will be read cold, mid-task, by an agent that has never seen the app.

1. Interview the repo, not the user

Answer these from the codebase and only ask the user what you cannot observe:

• Surface: what does a user actually touch? A web UI, a CLI/TUI, a desktop app, an API, a mobile app, a library? A repo can have several; pick the primary one and note the rest.
• Run: how does the app start locally? Prefer the repo's own documented dev command (package scripts, Makefile, README quickstart). Note ports, env vars, seed data, auth.
• Drive: how can an agent interact with it programmatically? Existing harnesses first — Playwright/Cypress specs, expect scripts, PTY helpers, curl-able endpoints, a debug port. Only then pick the OMP surface that fits: the eval tool's browser object for a web UI (it drives real Chromium tabs; read xd://eval/browser for the API), bash with pty for a CLI or TUI and a named bash service supervised through proc://<id> for a server that must stay up, plain HTTP with curl for services, and for a native macOS app the computer eval prelude when computer.enabled is on, otherwise cua-driver or peekaboo when installed. If none of these reaches the surface, say so in the generated skill and name the gap.
• Observe: what evidence can be captured? Screenshots, terminal transcripts, response bodies, logs, exit codes, DB state.
• Isolate: can two instances run side by side (ports, data dirs, profiles)? If not, say so in the generated skill: refusing to double-drive a shared instance beats corrupting the user's session.

If the checkout doesn't build or start as-is, fix that first (or report it precisely) before generating; a skill written against a broken base teaches wrong steps. When an irrelevant missing asset blocks startup (a static dir the API never serves, a sample config), the generated skill may create it, clearly marked as verification scaffolding, and remove it in cleanup.

2. Generate the skill

Write .omp/skills/verify-<app>/SKILL.md with YAML frontmatter (name: verify-<app>, the kebab-case directory name, and a description that names the app, the surface, and when to reach for it — without frontmatter the skill never registers) and these sections, each grounded in what the interview actually found (no placeholders left):