search-first

Researches existing libraries, MCP servers, and skills before writing custom code.

Updated Mar 25, 2026
One-click install
npx skills add https://github.com/Femad-6/my-skills --skill search-first-femad-6
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: search-first
Source: https://github.com/Femad-6/my-skills/tree/main/.github/skills/search-first
Command: npx skills add https://github.com/Femad-6/my-skills --skill search-first-femad-6

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Developers often reinvent the wheel by writing custom utilities when well-maintained open-source packages, MCP servers, or existing skills already solve the problem, leading to wasted effort and unnecessary maintenance burden. ## Core Features & Use Cases - Structured Research Workflow: Guides a five-phase process from need analysis through parallel search (npm/PyPI, MCP servers, GitHub), evaluation, decision, and implementation. - Decision Matrix: Provides clear adopt/extend/compose/build criteria based on candidate quality signals like maintenance status, license, and community. - Agent Integration: Coordinates with researcher, planner, and architect agents so technology decisions are informed by existing solutions. - Use Case: Before writing a custom HTTP retry wrapper, the skill searches npm and PyPI, finds got or httpx with built-in retry support, and recommends adopting them directly with zero custom code. ## Quick Start Ask the assistant to research existing tools and libraries for the functionality you are about to build before writing any custom code.

Frequently Asked Questions about search-first

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find existing libraries before writing custom code?

Search package registries like npm or PyPI for your problem domain, check for MCP servers and existing skills, then run GitHub code search for maintained open-source implementations. Evaluate candidates on functionality, maintenance, community, documentation, and license before deciding.

When should I adopt a library versus build a custom solution?

Adopt when an exact match exists that is well-maintained with a permissive license like MIT or Apache. Extend with a thin wrapper for partial matches, compose multiple small packages for weak matches, and only build custom when nothing suitable exists.

What is the search-first workflow for adding dependencies?

The workflow has five phases: define the needed functionality, run parallel searches across npm/PyPI, MCP servers, and GitHub, score candidates on quality signals, decide between adopt/extend/compose/build, then implement with minimal custom code.

Should I check for MCP servers before installing a package?

Yes, checking for MCP servers first can reveal that the capability is already available through an existing integration, eliminating the need for any new dependency. Check your MCP configuration and search available servers before adding packages.

What are the risks of skipping research before coding?

Skipping research leads to reinventing existing solutions, missing MCP-provided capabilities, over-wrapping libraries until they lose their benefits, and installing bloated dependencies for small features. A quick registry search prevents these anti-patterns.