skills.rest
liaochenjun
liaochenjunCOMMUNITY
Explore all skills by this author →
GitHub

reuse-oss

Searches and evaluates GitHub open-source projects before writing new feature code.

Updated Sep 11, 2026
Search Tags:#open source#code reuse#github#mcp server#license compliance#project evaluation#dependency assessment
One-click install
npx skills add https://github.com/liaochenjun/Cross-Border-E-commerce-Agent --skill reuse-oss-liaochenjun
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: reuse-oss
Source: https://github.com/liaochenjun/Cross-Border-E-commerce-Agent/tree/main/.claude/skills/reuse-oss
Command: npx skills add https://github.com/liaochenjun/Cross-Border-E-commerce-Agent --skill reuse-oss-liaochenjun

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Developers often waste effort reinventing functionality that mature open-source projects already provide. This Skill enforces a mandatory "reuse first" workflow so that new features start with searching, evaluating, and integrating existing GitHub projects, MCP servers, SDKs, or libraries instead of writing code from scratch. ## Core Features & Use Cases - Mandatory Search-First Workflow: Follows a fixed pipeline of requirement analysis, GitHub search, candidate comparison, user confirmation, integration, and verification before any custom development begins. - Structured Candidate Evaluation: Scores candidate projects against bundled checklists covering repository health, code quality, test quality, and dependency management, plus license compatibility checks against the repository's CC BY-NC-SA 4.0 terms. - Forced Reporting Gate: Requires reporting candidate projects, GitHub URLs, selection rationale, integration method, and license risks to the user for approval before any integration happens. - Use Case: When asked to add a web crawling feature, the agent searches GitHub via the GitHub MCP server, compares candidates like Crawl4AI against the evaluation checklists, reports findings, and only integrates after user confirmation. ## Quick Start Before implementing this new feature, search GitHub for mature open-source projects that already solve it and report the candidates for my approval.

Frequently Asked Questions about reuse-oss

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

FAQPage Schema
How do I find existing open-source projects before building a feature?▼

Search GitHub repositories using the GitHub MCP server, gh CLI, or the GitHub search API with keywords describing the needed capability. Compare candidates by stars, recent commits, issue activity, license, and documentation before choosing one to integrate.

How to evaluate whether a GitHub project is mature enough to integrate?▼

Check commit activity within the last 90 days, contributor count, issue response patterns, and release cadence. Then assess code quality via linting, test coverage above 70 percent, and dependency health using vulnerability scanners like pip-audit or npm audit.

What tools can search for MCP servers and plugins?▼

The claude-oracle-mcp server searches 17 registries covering over 15000 skills, plugins, and MCP servers with no API key required. The GitHub MCP server handles repository and code search, with gh CLI and WebSearch as fallbacks.

Does license compatibility matter when reusing open-source code?▼

Yes, license conflicts can block integration. Check for GPL copyleft, non-commercial restrictions, and redistribution requirements, and distinguish between using a tool as-is versus merging its code into a repository under a different license like CC BY-NC-SA 4.0.

When should I write custom code instead of reusing open source?▼

Only after confirming no suitable GitHub project exists or that existing candidates cannot meet the requirements. If the first candidate fails, keep searching alternatives rather than immediately switching to custom development.

Why does the GitHub MCP server show a tools fetch failed error?▼

The health check error MCP -32001 occurs because the large tools list response is slow over proxied networks, but actual tool calls still work normally. If real calls also time out, raise MCP_TIMEOUT or switch to a local stdio binary.

Related Skills

Explore More Software Engineering Skills→
edercnjedercnj

x-create-capability

Decompose a product into 3–7 capability artifacts with inherited RNF constraints.

Community
Advanced
knowz-ioknowz-io

kc-telemetry-setup

Configure Sentry and Application Insights telemetry sources with connectivity verification.

Official
Intermediate
AlexanderMattTurnerAlexanderMattTurner
64

vuln-research

Find, reproduce, and report security vulnerabilities against a stated threat model.

Official
Advanced
orqesaorqesa

build_vs_buy

Evaluate build versus buy options for software capabilities.

Official
Intermediate