king-context

Search indexed documentation and research corpora efficiently via the `.king-context/bin/kctx` CLI (list, search, read, topics, grep). Trigger when the user wants to query content that is already indexed locally — "search the docs / our research", "how to use library X", "what's the API for Y", "what do we have on topic Z", "find in indexed content". Do NOT trigger when the user asks to scrape a new doc site (use scraper-workflow) or to run a new open-web research pipeline (use king-research). Prefer this skill over raw web search whenever a local corpus may contain the answer.

Retrieve indexed project ADRs and architectural decision history through `.king-context/bin/kctx adr`. Use when the user asks why the project uses something, whether a decision already exists, what the current architecture decision is, what changed historically, or whether a planned code change conflicts with prior decisions.

Create and maintain project ADRs through `.king-context/bin/kctx adr`. Use when the user asks to create an ADR, register a decision, document an architecture choice, supersede an old decision, or record why the project changed direction.

Build an open-web research corpus on a topic using the king-research pipeline (generate → search → chunk → enrich → export) and auto-index it for later search. Trigger when the user asks to research / pesquisar / build a corpus / find sources / survey state-of-the-art on a general topic (papers, blog posts, discussions, comparisons). Do NOT trigger when the user points to a specific product documentation site — use the scraper-workflow skill for that. Do NOT trigger when the user wants to query already-indexed content — use the king-context skill (`kctx search`) for that.