npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@vovy-ai/context-engine

v0.2.1

Published

Vovy's Context Engine — deterministic, zero-LLM symbol search over a project via tree-sitter WASM (no native build, no network calls). Phase 1 of the semantic-search roadmap in docs/architecture.md.

Readme

@vovy-ai/context-engine

Deterministic, non-LLM symbol search for Vovy — a drop-in skill pack for vibe coding safely with AI coding assistants. Free forever, MIT, local-only.

You normally don't install this directly — it's what powers the search_codebase tool served by @vovy-ai/mcp-server once you run npx @vovy-ai/go install. It answers "where is X declared" / "what references X" / "is it safe to change X" questions directly, instead of an agent guessing a filename or grepping blind. No embeddings, no network calls, no LLM.

Two backends, chosen per project

  • typescript — used when TypeScript resolves from the project root, which is nearly always. Drives ts.LanguageService in-process against the project's own typescript, so symbols are scope- and type-aware and each reference names the declaration it resolves to. typescript is never a runtime dependency of this package; a project that has it pays nothing extra, and a project that doesn't gets the fallback.
  • tree-sitter — the fallback. Parses JS/TS/JSX/TSX with tree-sitter via web-tree-sitter's WASM runtime (no native build step). Matches whole identifier tokens, so never a substring inside a string or comment — unlike grep — but cannot tell two same-named symbols in different scopes apart.

Call backendKind(root) to find out which will answer. Every search_codebase response reports it too, so a name match is never mistaken for a resolved one.

What it exposes

  • getSymbolsOverview(filePath, root?) — declarations in one file and their members (class methods, interface members, object-literal methods), with line ranges and containing symbol.
  • findSymbol(name, root) — every declaration site matching name across a project.
  • findReferencingSymbols(name, root) — usage sites, excluding declarations. On the typescript backend each result carries the declaration it resolves to.
  • impactOf(name, root, maxDepth?) — transitive blast radius: usage sites of name, then usage sites of the declarations enclosing those, out to maxDepth hops (default 3), each tagged with its depth and caller. "What breaks if I change this."
  • searchPattern(query, root) — plain regex/text fallback for non-code content.
  • backendKind(root) / selectBackend(root) — which engine answers for this project.

Set VOVY_CONTEXT_BACKEND=tree-sitter to force the fallback (used by the benchmarks to A/B the two).

See docs/architecture.md's Context Engine section for the full design, the remaining honest limitations (JS/TS only, nothing persists across processes, no call graph), and scripts/eval-context-engine/RESULTS.md for a reproducible correctness benchmark scored against hand-verified ground truth.

License

MIT