← ClaudeAtlas

bookkeepinglisted

Universal knowledge engine — scores, promotes, and compounds knowledge across all sources into a permanent, query-able entity graph
broomva/skills · ★ 3 · AI & Automation · score 75
Install: claude install-skill broomva/skills
# bookkeeping — Universal Knowledge Engine The bookkeeping skill is **bstack primitive P6**: the universal knowledge bookkeeping layer that sits beneath every knowledge-producing workflow in the Broomva stack. It implements the LLM Wiki pattern (Karpathy): raw sources flow in, get scored, scatter into entity pages, deduplicate against the existing graph, and compound into synthesis notes. Every other skill that produces knowledge delegates its extraction and promotion phases here. --- ## When to Invoke - After any knowledge-gathering session: social engagement runs, research experiments, deep-dive sessions, conversation transcripts - When prompted with `/bookkeeping` or `bookkeeping run` - Automatically after the `social-intelligence` loop runs (Phase 2 — Knowledge Extraction — is fully delegated here) - Before creating synthesis notes or flagging blog post candidates - When asked to "extract knowledge from", "distill", "index", or "promote" any content - When entity pages are stale or lint errors are detected in the entity graph ## Reflexive Trigger Rule (binding on every agent in this workspace) Bookkeeping — and knowledge capture generally — is a reflex, not a request, **and never a question**. Agents must invoke `bookkeeping.py` and file graph-worthy material without being prompted in any of these situations: 1. **Before committing a feature/page that reads from the graph** — anything consuming `~/.config/bookkeeping/status.json`, `research/entities/`, or a snapsho