kuil09

Builds an API Goal Canvas from rough API, product, or workflow descriptions. Use when designing or reviewing APIs, restructuring requirements around actors, tasks, inputs, outputs, and goals, or detecting missing links, orphan records, and over-segmented request flows.

Git notes agent memory: canonical ref refs/notes/agent-memory for handoff; optional per-agent refs refs/notes/am-agents/<id> with aggregate for overlap signals.

A scientific thinking framework that treats every piece of information (user requests, agent responses, feedback, experiment interpretations) as a hypothesis to be verified through falsification-first experimentation. Use when asked to investigate, verify, debug, validate, test claims, test assumptions, analyze root causes, or when the task involves uncertainty, competing explanations, or non-obvious causality.

Use when integrating Toss Securities Open API for market data, holdings, order workflows, and approval-gated execution.

Convert informal conversations, meeting transcripts, planning notes, requirement drafts, and policy prose into a Lex Scripta document with parent statements, nested conditions, exceptions, action items, open items, assumptions, and risks. Use when natural language must be preserved but rewritten into an indentation-based structure that an AI or rules engine can interpret without guessing missing facts.

Use when receiving GitHub Actions results back into a local Codex or coding-agent loop through ci-local, Git notes, and Dumb Pipe.

Use when multiple AI agents need to share durable project context through a Git-backed remote repository. This skill standardizes how agents bootstrap a shared context repo, sync before reading or writing, record facts and decisions in Markdown, prepare branch-first updates, detect conflicts, and leave a clean handoff using only standard Git CLI and bundled Bash helpers.