Search vault notes by meaning using local embeddings
The English summary is being prepared. Showing the original description.
Run score Not tested
See how to installReview the instructions the agent will receive before you install.
Search the Obsidian vault by meaning, not just keywords. Uses local all-MiniLM-L6-v2 model — no API calls, no data leaves your machine.
# Search by meaning
node daemons/semantic-search/search-vault.js "your query here"
# Re-index (run after significant vault changes)
node daemons/semantic-search/embed-vault.js
# Re-index only changed files
node daemons/semantic-search/embed-vault.js --changed-only
cd daemons/semantic-search && npm install
node embed-vault.js # Initial index (~30 seconds)
If your vault also holds client work, deal notes, or anything under NDA, keep it out of the index:
cd daemons/semantic-search
cp index-deny.example.json index-deny.json
# then edit index-deny.json and list your own folders
Each entry in deny_prefixes is a vault-relative path prefix, matched case-insensitively and with either slash style. embed-vault.js drops matching files before they are embedded, so their text never reaches embeddings.json (which stores note excerpts in plaintext). search-vault.js re-checks the same list at query time, so an index built before you added a prefix still can't return one.
Only index-deny.example.json is tracked by git. Your index-deny.json is gitignored, because the names of your confidential folders are themselves confidential. With no index-deny.json at all, both scripts index and search everything and print a notice on stderr saying so. If the file exists but can't be read or parsed, they exit non-zero instead of falling back to an unfiltered index. Add a prefix the same session confidential material enters your vault, then re-run embed-vault.js: it also purges anything already indexed under a newly added prefix.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Track and surface sub-agent dispatch calibration ratios. Extract from JSONL transcript or read existing ledger. Per-agent clean/blocked/errored/partial trends.
The English summary is being prepared. Showing the original description.
Compose the AIgent's body state by reading existing signals. Lazy compute. Never polled. Reports the five pressures + recommended reflex.
The English summary is being prepared. Showing the original description.
Generate a structured delegation brief for any task
The English summary is being prepared. Showing the original description.
Quarterly skill that walks the skill-index against the actual SKILL.md files on disk, reports drift (skills indexed that no longer exist, skills on disk that aren't enrolled, skills whose descriptions have changed since enrollment). Hardens the auto-enrollment pipeline against silent rot.
The English summary is being prepared. Showing the original description.
When a [CADDY] hint fires and the principal is confused why, /caddy-explain walks the scoring transparently — which trigger matched, what points it earned, what the threshold was, what other skills were close. Pairs with the legibility thesis. Caddy's decisions should be inspectable on demand.
The English summary is being prepared. Showing the original description.
Enroll a skill into Caddy's index so it can be surfaced when the right prompt arrives. Use when a new skill is created, copied in, or when a skill needs updated triggers. Usage — "/caddy-enroll <path to SKILL.md>".
The English summary is being prepared. Showing the original description.