discoverlisted

# Discover Take a vague topic; emit structured forcing-questions (goal / constraints / scope / edge cases / acceptance) plus a concrete `sgc plan` follow-up. The user answers the questions inline, then carries the consolidated answer into `sgc plan --motivation`. **Core principle:** ambiguity in requirements multiplies into bugs in code. Eliminate it before writing a single line. ## When to Use - User says "I want to build..." but details are vague - Requirements have obvious gaps or contradictions - Multiple valid interpretations exist for the same request - Before `/plan` when the task is L2+ and scope is unclear ## Permission | Directory | Access | |-----------|--------| | decisions | — | | progress | R | | solutions | — | | reviews | — | Plus `spawn:clarifier.*`. No writes — `discover` cannot mutate `.sgc/` state beyond the spawn audit trail under `progress/agent-prompts/` and `agent-results/`. ## Routing - **Behavior**: [`src/commands/discover.ts`](../../../../src/commands/discover.ts) (`runDiscover`) - **Agent**: [`src/dispatcher/agents/clarifier-discover.ts`](../../../../src/dispatcher/agents/clarifier-discover.ts) — heuristic stub keys off auth / data / ui / perf / api keywords to tune the question set - **Contract**: `clarifier.discover` manifest in [`contracts/sgc-capabilities.yaml`](../../../../contracts/sgc-capabilities.yaml) ## Execution When this skill is invoked, dispatch to the sgc CLI: ```bash bun src/sgc.ts discover "$ARGUMENTS" ``` Output is st