discoverlisted

# Discover — Conversational <!-- BUDGET_EXCEPTION: Entry-point conversational skill documents all three depth tiers (Light / Medium / Deep) in one body. Compacting below ~2,500 tokens removes structural meaning. Auto-downgrade routes simple invocations to Light; only Medium/Deep depth loads the procedural references. --> Transform vague ideas into shared clarity through adaptive conversation. Capability metadata + load map: [`routing.yaml`](routing.yaml). Methodology (conversation-IS-the-alignment, blunt-peer stance, equal-weight rule, resolution-exit-not-patience-exit) + when NOT to use: [`references/playbook.md`](references/playbook.md). **Core philosophy:** Just talk with your agent. Close the gap between stated requirements and true needs through conversation — not documents, formal phases, or plan mode. **Core question:** What would we silently get wrong if we just started building? ## How It Works You describe what you want → agent scans context + assesses complexity silently → adaptive questions begin → conversation continues to mutual clarity → build directly or save a spec. No plan mode. No pipeline stages. No mandatory artifacts. **The conversation IS the alignment.** ## Before Starting Apply [`references/_shared/before-starting-check.md`](references/_shared/before-starting-check.md). Then: - **Mode resolution** ([`references/_shared/mode-resolver.md`](references/_shared/mode-resolver.md)). `budget: fast`; Adaptive Depth auto-calibrates Light / Medium / Dee