Explore AI Agent Skills & Claude Prompts

Use this skill when the user wants to find something specific in their Second Brain — any search query, lookup, or "where did I write about X?" request. This covers searching across vault files, journal entries, concepts, action items, and the knowledge graph simultaneously, returning ranked results with graph-adjacent connections. Also use when the user wants to locate a note, recall when they discussed a topic, or check if they already captured something. Prefer this skill for any retrieval request, even vague ones. Distinguished from brain-trace (which follows a concept's evolution over time, not a point-in-time search), brain-resources (which catalogs the entire knowledge base rather than answering a specific query), and simple file reads (this searches across vault, DB, and knowledge graph together).

brain-context-load

Use this skill when the user wants to orient a fresh session with their Second Brain state — any request that signals "get up to speed on who I am and what I'm doing" without needing a daily note or morning briefing. This includes loading context at the start of a session, asking to be caught up, wanting to know current projects and priorities, or requesting a status snapshot. Reads identity files (ICOR, Values, Active Projects), queries SQLite for recent journal summaries, pending actions, attention flags, and active concepts. Distinguished from the morning review (which also loads context but additionally creates today's daily note and generates a morning briefing) — use context-load when you just need awareness, use brain-today when you're starting your day.

pr-review

Use this skill when reviewing code changes before committing to main — any request to review a diff, audit recent changes, or validate that modifications follow project conventions. This includes blast radius analysis (which modules import the changed code), security scanning (SQL injection in SQLite queries, credential exposure, unsafe eval), test coverage delta with pytest, breaking change detection (DB schema migrations, config key removals, ICOR hierarchy changes), and performance impact assessment (N+1 queries, unbounded loops, embedding recomputation). Produces a structured review with MUST FIX, SHOULD FIX, and SUGGESTIONS tiers plus a project-specific checklist. Distinguished from security-audit (which scans an entire directory for installation safety, not a commit diff) and from the code-simplifier agent (which refactors for clarity, not correctness or security).

schedule Updated 2 months ago

brain-sync-notion

Use this skill when the user wants to synchronize data between their local vault/SQLite and the Notion workspace — pushing local changes to Notion or pulling remote updates back. This covers any request to sync tags, tasks, projects, goals, journal entries, concepts, or people with Notion, whether full or selective. Also use when the user mentions data being out of date between systems, wants to push action items to Notion Tasks, or pull project status changes back locally. Distinguished from brain-refresh-dashboard (which only recalculates metrics and updates the cockpit, not raw entity data) and from one-off Notion API queries (this runs the full bidirectional sync pipeline with conflict resolution and state tracking).

brain-emerge

Use this skill when the user wants to discover hidden patterns, implicit themes, or unnamed directions lurking across their notes — any request about "what patterns do you see," surfacing unconscious priorities, finding recurring questions, or synthesizing scattered observations into coherent insights. This includes detecting sentiment shifts, thematic clusters, converging interests, and conspicuous absences. Distinguished from brain-drift (which compares behavior against stated goals rather than discovering new patterns) and brain-graduate (which promotes already-identified recurring themes into concept notes rather than discovering them for the first time).