After this, your AI stops generating plausible-looking code against assumptions you never made — and starts asking the right questions before any code gets written. Who this is for Solo developers and small teams who use Claude Code, Cursor, or Codex daily. You feel the friction: the AI dives in too fast, generates 400 lines against guesses, you spend 30 minutes correcting drift. You want the assistant to behave like the colleague who's been here before — pausing for the right question instead of plowing ahead. Before — default Claude, prompt: "Add Stripe subscription billing" Streams 400 lines of code immediately: Stripe SDK setup, route handlers, billing cycles, three plan tiers. Half of it built against assumptions you never made (monthly vs. annual? self-serve cancellation? proration rules?). You spend the next hour ripping out the 60% you didn't want. After — same prompt, with Solo Dev Set installed Q: What's the billing model? (a) Subscription [recommended] (b) Usage-based (c) Hybrid > a Q: How many tiers? (a) Two (b) Three [recommended] (c) Four+ > b [six more questions — annual option, free trial, proration, cancellation flow, dunning, invoicing. Each with a recommended answer. Three minutes.] Now writes the Stripe integration against the decisions you actually made. No wasted code. No rip-out hour. What's inside decisions-grill — interrogates the plan one question at a time, with a recommended answer for each, until every load-bearing decision is resolved tests-first — forces a failing test before any production code; bugs caught at write-time root-cause-debug — replaces guess-and-check thrashing with a four-phase root-cause process before any code changes deepen-modules — surfaces only the refactors that earn their keep; small stable interfaces over loud rewrites send-for-review — prepares the changes summary, evidence, and risk notes BEFORE you open the PR terse-mode — cuts AI output roughly 75% with no loss of technical substance Three ways to use this Paste it (any AI, 30 seconds, no install)Open the .md file, copy contents, paste into Claude / ChatGPT / Cursor / Perplexity with: "Read this and apply it whenever the task it describes comes up in our conversation." Done. Install it (Claude Code / Cursor users)Drop the unzipped folder into ~/.claude/skills/ (Claude Code), .cursor/skills/ (Cursor), or your IDE's equivalent. Auto-loads next session. New to AI assistants? Get Claude free at claude.ai, then use Option 1. €24 — six skills · sum of parts €42
