second-brain

Operate and maintain the user's local second-brain — health checks, re-embedding, corpus backfills, summary mirror sync, derived-link rebuild, People Hub regeneration, and embedder swaps. Use this skill when the user asks whether the brain is healthy, asks to rebuild embeddings, asks to resync summaries to the wiki, asks to regenerate the people pages, asks to refresh derived links or directory, or reports a brain CLI error that suggests an ops-level repair. Covers brain doctor / status / analyze / reembed / init / backfill / vault sync-summaries / vault relink-derived / vault directory refresh. MANDATORY TRIGGERS: brain doctor, is my brain healthy, brain status, rebuild embeddings, reembed, brain reembed, backfill the brain, backfill tags, backfill search, sync summaries, sync summaries to wiki, regenerate people pages, refresh derived links, refresh directory, relink, switch embedder, ollama down, postgres down, brain init, brain analyze, analyze tables, refresh planner stats, chunks stats warn.

Show open action items the user has captured in their local second-brain, and orchestrate the deferred-fetch flow that ingests action items from Krisp meetings into the brain. Use this skill whenever the user asks what's on their plate, what action items they owe, what follow-ups came out of recent meetings, or asks to pull Krisp action items. Backed by `brain todo` (parser/view) and `brain enrich --krisp-action-items` (deferred-fetch handoff that the agent completes via the Krisp MCP). MANDATORY TRIGGERS: what's on my plate, my action items, my todos, my TODOs, open action items, follow-ups, what do I owe, what do I need to do, krisp action items, pull action items, ingest krisp action items, brain todo, action items from my meetings, what are my next steps.

Answer questions or write content using the user's local "second brain" — a hybrid-search knowledge base of career artifacts, meeting transcripts, Slack threads, Gmail, and authored vault notes. Use for lookups about the user's own history, quotes, facts, and drafting in their voice. For themes / patterns / connections across interactions, use `brain-graph` instead. MANDATORY TRIGGERS: ask my brain, second brain, what did I say, my conversations, my notes about, write in my voice, draft like me, my career, my interview prep, summarize my emails about, what did I tell, did I ever say, remind me about my, pull from my brain, search my notes for, do I have anything written about.

Surface and codify tacit knowledge from the user's local second-brain — knowledge implied everywhere in their transcripts and notes but written nowhere. Use this skill when the user wants to discover what they know implicitly, fill in knowledge gaps, or turn unwritten rules into authored vault notes. Backs `brain elicit` (interactive draft-then-correct loop, needs Ollama) and `brain elicit list` (Ollama-free gap queue view). For plain document lookup or Q&A use `consult-brain`; for themes and patterns across interactions use `brain-graph`. MANDATORY TRIGGERS: elicit my tacit knowledge, what do I know that I haven't written down, what haven't I written down, surface my knowledge gaps, surface my tacit knowledge, what implicit knowledge, draft the unwritten rules, what rules am I not writing down, interview me about, what do I keep referencing without writing, brain elicit, knowledge elicitation, codify what I know, fill in my knowledge gaps, what's in my head that isn't in my notes, tacit knowledge, gap in my

Graph retrieval over the user's local "second brain" — an entity-centric GraphRAG layer (Apache AGE) alongside hybrid search. Use for THEMES, PATTERNS, or CONNECTIONS across interactions, or to ENUMERATE the graph (people / orgs / projects / topics). For single-doc lookups, quotes, or voice writing, use `consult-brain` instead. MANDATORY TRIGGERS: themes with, themes in my conversations, patterns across, what connects, what keeps coming up, recurring themes, recurring topics, how has my thinking evolved, map out my, who is related to, what is related to, the bigger picture across, graph of my, cluster my notes, overall themes in my brain, what orgs are in my brain, what people are in my brain, list all projects, what topics are in my brain, what entities, how big is my graph, what's in my graph, graph overview, graph stats.

Create and edit vault-tier notes in the user's local second-brain via the `brain` CLI. Use this skill when the user wants to start a new note, open or create today's daily entry, rename an existing note (with reference refactoring), edit a note's frontmatter or body, add/remove tags, ask the LLM to propose tags, or quarantine / un-quarantine a note as draft. Covers brain note new / brain daily / brain note rename / brain edit / brain tag / brain mark-draft / brain mark-published. MANDATORY TRIGGERS: new note, create a note, start a note, scaffold a note, daily note, today's daily, daily entry, rename this note, rename my note, edit this doc, edit this note, tag this, untag this, auto-tag, propose tags, mark as draft, mark as published, quarantine this note, unhide this note.

Add new content to the user's local second-brain knowledge base using the `brain` CLI's ingestion commands. Use this skill whenever the user asks to ingest, save, pull, or capture content into the brain — including files (PDFs / DOCX / Markdown / TXT), directories of files, Gmail threads, Krisp meeting transcripts, Slack threads, or arbitrary stdin snippets. For Krisp/Slack the brain does NOT have a direct integration; instead orchestrate the relevant MCP server and pipe the fetched content into `brain ingest-stdin`. Always pick a sensible scope flag — never bulk-ingest the Gmail inbox. MANDATORY TRIGGERS: ingest, ingest last week, ingest this file, ingest this directory, pull this email, pull last week's Krisp calls, save this Slack thread, save this transcript, add this to my brain, capture this in my brain, import this into the brain, brain ingest, brain ingest-stdin, brain ingest-gmail.