zvec

A command-line tool for finding relevant project documents by meaning rather than exact words. It searches an indexed knowledge base, such as documentation, regulations, and business-process notes.

In plain words
What is it for?
Use it to locate related documentation, regulations, process descriptions, or context before changing project configuration; use ordinary grep for code and configuration files.
Why use it?
A normal text search may miss documents that use different wording, while searching source code with this tool is outside its intended scope.

Cursor rule for Cursor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/alexqem/zvec-tool/zvec
Clone the repo
git clone --depth 1 https://github.com/alexqem/zvec-tool

Made for: Cursor.

Per session 0 Nothing until a file matches its globs; then the whole rule loads.
When invoked 765 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00765
Opus 5 $0.00000 $0.00382
Sonnet 5 $0.00000 $0.00153
Haiku 4.5 $0.00000 $0.00076

Measured yesterday against content hash ba6b422b137d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

zvec scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.cursor/rules/zvec.mdc · 82 lines

How it starts

The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.

zvec — semantic search skill

zvec-tool is a CLI for semantic search over the project knowledge base. It turns documents into a vector index and retrieves relevant fragments by meaning. Use it to find docs, regulations, and business-process notes before doing broad greps.

  • Model: intfloat/multilingual-e5-small (384-dim, multilingual)
  • Indexed scope: the folder you pass to index (only *.md by default; configure via ZVEC_EXTENSIONS)
  • No global DB lock: withCollection opens/closes the DB per call, so search never blocks index

When to use

Call zvec search when you need to:

  • find how a business process is described/documented
  • locate documentation for a specific entity (справочник/документ/регистр)
  • find related regulations or concepts
  • get context before changing project configuration

Do not use zvec to search source code or config files — it only indexes the document folder you configured. For code, run a normal grep instead.

Search command

Run from the project root (the parent of zvec-tool/):

node zvec-tool/zvec.mjs search "<query in RU or EN>" --top 8
Flag Description
--top N Number of results (default: 8)
--json Machine-readable JSON output

Examples

node zvec-tool/zvec.mjs search "учёт товаров на складе" --top 5
node zvec-tool/zvec.mjs search "регистр накопления остатки" --top 8
node zvec-tool/zvec.mjs search "порядок закрытия месяца" --top 6
node zvec-tool/zvec.mjs search "документ поступление" --json

Each result contains: the file path, a score (cosine similarity, > 0.7 is strong), and a 220-char preview. Results are sorted by descending score.

Rebuild the index

# Path is required — point it at your documents folder
node zvec-tool/zvec.mjs index ./docs
# or: node zvec-tool/zvec.mjs index --root ./docs
# or set ZVEC_INDEX_ROOT=./docs

Status & diagnostics

node zvec-tool/zvec.mjs status     # index status: docCount, active version
node zvec-tool/zvec.mjs doctor     # detailed diagnostics (probes DB, shows lock holder)

Read the full file on GitHub · 82 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 82 lines · 0 tokens per session scan A ba6b422b137d

Subscribe to this mod's changes

zvec is a cursor rule published in the GitHub repository alexqem/zvec-tool (2 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 765 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.