Drop notes in. The pipeline sorts, indexes, and surfaces what matters. Nothing fancy. Just a system that works.
vqi
vqq
vqr
Daily digest runs at 6 PM IST
Last digest: June 6, 2026. Next auto-run: today at 6 PM IST.
Three commands run the whole system. They live in ~/.zshrc.
Re-index the vault. Updates search + stats. Run after dropping new raw files.
Query Honcho for cross-domain synthesis. Finds patterns across disconnected notes.
Generate today's digest. Shows orphans, stale notes, and vault health. Auto-saves to wiki/.
A cron job runs the digest every evening. No manual trigger needed unless you want it early.
# ~/.zshrc
alias vqi='cd ~/Documents/hermes-vault && python3 scripts/vault-indexer.py'
alias vqq='cd ~/Documents/hermes-vault && python3 scripts/vault-query.py'
alias vqr='cd ~/Documents/hermes-vault && python3 scripts/vault-review.py \
--index --days 1 --title "Daily Vault Digest" && \
cat wiki/daily-review-$(date +%Y-%m-%d).md'
The daily cron at 6 PM IST runs the same sequence: index : review : save. Output lands in wiki/daily-review-YYYY-MM-DD.md. Read it or ignore it : no task spam, no Todoist bridge.
Every session gets stored in Honcho's RAG stack. Cross-session queries surface past decisions, preferences, and stale project threads. The vqq alias queries this for synthesis, connecting notes from different days into coherent themes.