blindspotlisted

# Blindspot Pass — Blind-Spot Investigation Before implementation, find the **gap between the map (plan/prompt) and the territory (actual codebase, domain, constraints)**. Origin: Blindspot Pass — see skills/loop/references/talk-source.md ## Iron Rules 1. **Do not implement yet.** No code changes while this skill is active. 2. End goal of the investigation: help the user **write a better prompt**. ## Procedure 1. Parse task description and context sources from `$ARGUMENTS`. Empty → the task currently under discussion; if there is none, ask one question first. If sources are given (git history, docs, specific modules, etc.), investigate those first. **Before investigating, ask once** (one AskUserQuestion, or one message) for three lists: ① what the user is already sure of, ② what they are assuming, ③ where they do not know how to judge. ① narrows the scope, but ask how recent that certainty is; ② becomes the **first** thing to check, not something to skip; ③ points at the **teach-me** skill. A skip means a full investigation — never block on it. 2. Scope codebase-wide or large → spawn the `unknowns:unknowns-scout` agent with the Agent tool. Hand it a packet: the user's original prompt verbatim, the context sources, the target area and its entry points, and what to return (finding table + improved prompt draft); map its rows onto the card kinds below. Scope is a few files → investigate directly with Read/Grep/Glob. No Agent tool in this