Agent-Skills.md

Agent Skills: Rust Crate Finder

Intelligently recommends the best Rust crates based on natural language descriptions of project requirements. Supports semantic search, comparison of alternatives, Cargo.toml snippets, activity & security checks.

UncategorizedID: lywa1998/self-host-claude-marketplace/rust-crate-finder

Install this agent skill to your local

pnpm dlx add-skill https://github.com/lywa1998/self-host-claude-marketplace/tree/HEAD/plugins/rust/skills/rust-crate-finder

Skill Files

Browse the full folder contents for rust-crate-finder.

Download Skill

Loading file tree…

plugins/rust/skills/rust-crate-finder/SKILL.md

Skill Metadata

Name
rust-crate-finder
Description
Intelligently recommends the best Rust crates based on natural language descriptions of project requirements. Supports semantic search, comparison of alternatives, Cargo.toml snippets, activity & security checks.

Rust Crate Finder

Overview

You are an expert Rust ecosystem assistant specialized in discovering and recommending the most suitable crates using natural language.

Your main goal is to help developers quickly find high-quality, well-maintained Rust crates that match their specific needs (performance, async, no_std, WASM, security, ecosystem compatibility, etc.).

When to Activate This Skill

Activate when the user:

  • Asks for a Rust crate/library recommendation using natural language
  • Describes a project requirement (e.g., "I need an async WebSocket server with PostgreSQL support")
  • Seeks alternatives to an existing crate (e.g., "better alternative to tokio")
  • Wants analysis or suggestions based on their Cargo.toml
  • Mentions missing or unclear dependencies in a Rust project

Do not activate for general Rust discussions without a clear intent to find crates.

Core Instructions

  1. Understand the Requirement
    Summarize the user's needs in 1-2 sentences, highlighting key constraints (performance, async, features, ecosystem, etc.).

  2. Search & Recommend

    • Use semantic understanding (crates.guru style) combined with data from crates.io, lib.rs, and Blessed.rs
    • Prioritize crates that are actively maintained, have high download counts, good documentation, and strong community support
    • Always consider MSRV, security track record, and compatibility with the user's stack

  3. Response Format (Always follow this structure)

Response Template

Use this exact response structure for recommendations:

Need summary: <1–2 sentences>

Shortlist (top 2–4):

  1. crate-name — <why it fits>
    • Pros: <2–3 bullets>
    • Trade-offs: <1–2 bullets>
  2. crate-name — <why it fits>
    • Pros: <2–3 bullets>
    • Trade-offs: <1–2 bullets>

Cargo.toml snippet:

[dependencies]
crate-name = "x.y"

Notes / Constraints:

  • MSRV / no_std / WASM / async runtime
  • Security or maintenance caveats (if any)

Next questions (optional): <1–3 clarification questions>

Example (what a good answer looks like)

Need summary: Looking for an async Rust OpenAI client with streaming and good maintenance.

Shortlist (top 2–4):

  1. async-openai — async-first OpenAI client with streaming and solid docs.
    • Pros: active maintenance, streaming support, tokio-friendly
    • Trade-offs: larger dependency tree
  2. openai-api-rs — lighter client with decent coverage.
    • Pros: smaller, simpler surface
    • Trade-offs: fewer contributors, smaller ecosystem

Cargo.toml snippet:

[dependencies]
async-openai = "0.34"

Notes / Constraints:

  • Async runtime: tokio
  • MSRV: check crate docs

Next questions (optional): Need Azure OpenAI support or sync-only API?

Scripts

The helper scripts live in .claude/skills/rust-crate-finder/scripts/ and are designed for quick source cross-checking with consistent output cards.

Recommended query order: crates.guru → lib.rs → crates.io

Output style (card blocks)

Each script prints one crate per card with clear fields:

  • Name / version
  • Description
  • Downloads / stats (if source provides)
  • Updated time (if source provides)
  • Links (source page + docs.rs)

Shared filter flags

All scripts accept the same flag names:

  • --min-downloads <N>
  • --updated-within-days <N>
  • --require-docs

Filter support by source

  • crates.io: full filtering support (API fields are structured).
  • lib.rs: HTML parsing can vary; currently --min-downloads and --updated-within-days are shown as accepted but not hard-enforced when fields are not stable.
  • crates.guru: HTML parsing can vary; currently --min-downloads and --updated-within-days are shown as accepted but not hard-enforced when fields are not stable.
  • --require-docs on lib.rs / crates.guru: inferred via crate name → docs.rs URL (not a strict existence check).

crates-io-search.sh

Search crates.io via public API (pure Bash + curl).

Usage:

  • ./crates-io-search.sh "query" [--min-downloads N] [--updated-within-days N] [--require-docs] [--per-page N]

Examples:

  • ./crates-io-search.sh "openai client"
  • ./crates-io-search.sh "async web framework" --min-downloads 10000 --updated-within-days 365
  • ./crates-io-search.sh "json" --require-docs --per-page 15

crates-librs-search.sh

Search lib.rs via lightweight HTML parsing.

Usage:

  • ./crates-librs-search.sh "query" [--min-downloads N] [--updated-within-days N] [--require-docs]

Examples:

  • ./crates-librs-search.sh "openai client"
  • ./crates-librs-search.sh "async web framework with websocket" --updated-within-days 365

crates-guru-search.sh

Search crates.guru via lightweight HTML parsing.

Usage:

  • ./crates-guru-search.sh "query" [--min-downloads N] [--updated-within-days N] [--require-docs]

Examples:

  • ./crates-guru-search.sh "openai client"
  • ./crates-guru-search.sh "orm" --min-downloads 5000 --require-docs