← ClaudeAtlas

map-mine-movelisted

Rapidly gain working expertise in an unfamiliar domain using a three-phase system — map the landscape, mine the failure points, move to action. Use when the user needs to get up to speed on a new field, prepare for a meeting outside their expertise, evaluate vendors or partners, or make a high-stakes decision in an area where they are not the specialist. Triggers include "get me up to speed on X", "I have a meeting about Y tomorrow and I know nothing", "help me evaluate this vendor/proposal", "what am I missing about Z", or any request to go from zero to dangerous in a domain. Not for casual curiosity, quick factual lookups, or domains the user already works in.
yorilavi/map-mine-move · ★ 0 · AI & Automation · score 65
Install: claude install-skill yorilavi/map-mine-move
# Map, Mine, Move — Rapid Domain Expertise Skill > **v1.0.0** · by Yori Lavi · updated 2026-08-19 · see [CHANGELOG](./CHANGELOG.md) > Requires: Claude Code Guide the user through a three-phase process to go from zero to dangerous in any unfamiliar domain. The goal is not to become a specialist, but to know the right questions, traps, and first moves. **CRITICAL**: All three phases MUST happen in the same conversation to build compounding context. Do not suggest starting a new chat between phases. **CRITICAL**: Deliver ONE phase per turn. Each phase ends with a question to the user — stop there and wait for their answer. Do not run Phase 2 until they have responded to Phase 1, and do not run Phase 3 until they have responded to Phase 2. The user's reaction to each phase is the context the next phase is built on; skipping the handoff throws away the compounding this skill exists to create. ## Before Starting Ask the user: 1. **What domain or situation** do you need to get up to speed on? 2. **What's your role** in this situation? (e.g., business owner, manager, investor, new hire) 3. **What's the context?** (e.g., hiring a vendor, entering a new market, evaluating a proposal) 4. **What's the decision and the stake?** What are you about to commit to, by when, and what does it cost if it goes wrong? Encourage the user to provide maximum context — the more detail, the better the output. Remind them not to self-edit; messy context is better than polished but incomplete conte