agentleFS
LucasCooperBeyLucasCooperBey/claude-skills/skills/package-research/SKILL.md

package-research

Skill1 starsChanged 5 months agoMITView on GitHub

Researches packages, dependencies, and APIs before integration. Leverages package-search MCP to ground responses in real source code.

Files
SKILL.md

Package Research Workflow

Ground your package research in real source code using the Package Search MCP.

When to Use This Skill

  • Evaluating packages before adding as dependencies
  • Finding correct API usage patterns
  • Understanding package internals
  • Comparing packages for decision-making
  • Validating that a package supports needed features

Available Tools

1. Semantic Search (Discovery)

package_search_hybrid(
  query="how to validate schemas",
  package="zod",
  registry="npm"
)

Best for: General questions, exploring functionality, "how do I...?"

2. Exact Match (Precision)

package_search_grep(
  pattern="export.*function",
  package="drizzle-orm",
  registry="npm"
)

Best for: Finding specific patterns, function signatures, imports

3. Read Source (Deep Dive)

package_search_read_file(
  file_path="src/index.ts",
  package="hono",
  registry="npm"
)

Best for: Understanding implementation, reading exports, studying patterns

Supported Registries

RegistryLanguages
npmJavaScript, TypeScript
pypiPython
goGolang
crates.ioRust
rubygemsRuby
terraformInfrastructure as Code

Research Workflow

Phase 1: Discovery

# What does this package offer?
package_search_hybrid("main features exports", package="<name>", registry="<reg>")

Phase 2: API Patterns

# How do I use the main API?
package_search_grep("export.*class|export.*function", package="<name>", registry="<reg>")

Phase 3: Implementation Details

# How does this specific feature work?
package_search_read_file(file_path="src/core.ts", package="<name>", registry="<reg>")

Example Research Sessions

Evaluating a New ORM

# 1. What query patterns does it support?
package_search_hybrid("query builder select where join", package="kysely", registry="npm")

# 2. How does it handle migrations?
package_search_grep("migration|migrate", package="kysely", registry="npm")

# 3. Read the migration implementation
package_search_read_file(file_path="src/migration/migrator.ts", package="kysely", registry="npm")

Finding Authentication Patterns

# 1. How does this auth library handle JWT?
package_search_hybrid("jwt token verify decode", package="jose", registry="npm")

# 2. Find the exact function signatures
package_search_grep("export.*verify|export.*sign", package="jose", registry="npm")

Python Package Research

# 1. How does requests handle retries?
package_search_hybrid("retry backoff session", package="requests", registry="pypi")

# 2. Find retry adapter implementation
package_search_grep("class.*Retry|def.*retry", package="urllib3", registry="pypi")

Why Use This Over Web Search?

  • Source of truth: Real source code, not outdated blog posts
  • No hallucinations: Grounded in actual implementations
  • Version-specific: See how the package actually works
  • Deep dive: Access internal implementation details
  • Cross-language: Same workflow for all supported registries

Integration with Other Skills

  • api-integration: Research package patterns before implementing clients
  • mcp-builder: Study MCP SDK internals before building servers
  • test-driven-development: Find test patterns in packages you use

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.