← ClaudeAtlas

prompt-masterlisted

Compile a raw idea (feature, refactor, migration, bug hunt, research) into a sequenced pack of copy-paste-ready prompts for Opus/Sonnet-class models — one prompt per development phase (research → spec → plan → implement → verify), each self-contained, grounded in the project's real facts, and chained through artifact files so the methodology survives fresh sessions and less self-driving models. TRIGGER when: the user wants prompts to run later or elsewhere ("generate prompts for this idea", "prompt pack", "how would I drive Opus/Sonnet through this", "разложи идею на промпты"), or wants a phased prompt plan instead of doing the work now; ALSO when the input is a whole surface rather than one change ("audit this whole area", "review every screen", "map all of X and tell me what to fix") — that is the program pass, and it is authored here, not by /prepare. DO NOT TRIGGER when: the user wants the work done in THIS session (route via /which-skill or run the pipeline skills directly), wants a spec (/analyst), a pl
mik2win/foureyes · ★ 2 · AI & Automation · score 79
Install: claude install-skill mik2win/foureyes
# Prompt Master Turn an idea into a **prompt pack**: an ordered sequence of prompts, one per development phase, that a user can paste into Opus/Sonnet-class sessions and get the same disciplined flow this kit runs natively — research before deciding, spec before planning, plan before code, verify before done. You do NOT do the work itself; you write the prompts that make another model do it well. Idea: `$ARGUMENTS` (if empty, ask what to compile, then stop until answered). Why this exists: stronger models self-drive the methodology; weaker ones need it **written into every prompt** — explicit method steps, guardrails, an output contract, and a stop condition. Each generated prompt must be runnable in a *fresh session* with nothing but the pack and the artifact files earlier prompts produced. **Two shapes come out of here.** The usual one is the phase chain above — one coherent change, one prompt per phase. The other is the **program pass** (tier 1): a whole-surface discovery sweep that ends in evidence files and *cards*, which `/prepare` then turns into plans. Same skill, different pipeline — see *Program packs* below and `references/pipelines.md` § 7. Deciding which one applies is Phase 1's job. ## Phase 0 — Ground (optional, don't block) Works with or without a kit profile: - If `.claude/PROJECT.md` exists, read it and hold: **Domain** (vocabulary, roles), **Stack & Architecture** (layers, dependency direction), **commands** (test, lint, run), and the **Plans loca