// Systematic approach for gathering authoritative, version-accurate documentation. Claude invokes this skill when research is needed before implementation. Ensures truth over speed while achieving both.
Systematic methodology for identifying, capturing, and documenting reusable patterns from implementations. Enables automatic learning and knowledge-core.md updates. Claude invokes this after successful implementations to preserve institutional knowledge.
Active context curation to fight context rot. Curates what goes into limited context window from constantly evolving information universe. 39% improvement, 84% token reduction.
Systematic validation methodology for ResearchPacks and Implementation Plans. Provides scoring rubrics and quality gates to ensure outputs meet standards before proceeding to next phase. Prevents garbage-in-garbage-out scenarios.
Systematic approach for creating minimal-change, reversible implementation plans. Claude invokes this skill when transforming requirements/research into executable blueprints. Emphasizes simplicity, safety, and clear verification steps.
| name | research-methodology |
| description | Systematic approach for gathering authoritative, version-accurate documentation. Claude invokes this skill when research is needed before implementation. Ensures truth over speed while achieving both. |
| auto_invoke | true |
| tags | ["research","documentation","verification","truth"] |
This skill provides a systematic methodology for conducting rapid, accurate documentation research to ground implementations in truth rather than potentially stale LLM memory.
Claude will automatically invoke this skill when:
Objectives:
Actions:
Parse request - Extract library/API names mentioned
Version detection - Check project dependency files:
package.json → Node.js projectsrequirements.txt, pyproject.toml, Pipfile → Pythongo.mod → GoCargo.toml → Rustbuild.gradle, pom.xml → Java*.csproj → C#/.NETpubspec.yaml → Dart/Fluttercomposer.json → PHPContext gathering - Note runtime, platform, existing dependencies
Output:
Target: [library-name]
Current Version: [X.Y.Z] (detected from [file])
Platform: [Node.js 20.x / Python 3.11 / etc.]
If unclear: Ask ONE specific clarifying question rather than proceed with assumptions
Source hierarchy (in order of preference):
Official documentation (PRIMARY)
Official migration/upgrade guides (if version change)
Official release notes/changelog (for version-specific info)
Official GitHub repository (if docs sparse)
Avoid (unless no alternatives):
Retrieval strategy:
1. Try context7 system (if available)
└─ Fastest, curated, version-aware docs
2. Use WebFetch on known official doc URLs
└─ Direct fetch from source
3. If URL unknown, use WebSearch
Query format: "[library name] [version] official documentation"
└─ Find the official site first, then fetch
4. Extract only relevant sections
└─ Don't download entire docs, target specific info needed
What to extract:
Anti-stagnation:
For every piece of information extracted:
API: someFunction(param1: Type): ReturnType
Source: official-docs.com/api-reference/someFunction [version X.Y.Z]
Verification checklist:
Confidence levels:
Report confidence: Always indicate confidence level in research output
Deliver in ResearchPack format:
# 📋 ResearchPack: [Library Name]
## Quick Reference
- Library: [name] v[X.Y.Z]
- Official Docs: [URL]
- Confidence: [HIGH/MEDIUM/LOW]
## Key APIs
[Concise list with signatures and sources]
## Setup
[Essential steps only, not every option]
## Gotchas
[Version-specific issues to avoid]
## Example
[Minimal working code with source link]
## Implementation Checklist
[Files to modify, steps in order]
## Sources
[Numbered list with URLs and versions]
## Open Questions
[Any decisions/clarifications needed]
Keep it concise:
Before delivering research:
Completeness check:
Accuracy check:
Usefulness check:
❌ Don't:
✅ Do:
Scenario: User says "Add Redis caching to the product service"
Claude (using this skill):
[Internal: This requires research on Redis client library]
🔍 Activating research-methodology skill
📦 Detecting Redis version from package.json... Found redis@4.6.0
🌐 Fetching official docs from redis.io...
[Follows methodology: assess → prioritize → retrieve → verify → output]
# 📋 ResearchPack: Redis Client (Node.js)
## Quick Reference
- Library: redis v4.6.0
- Official Docs: https://redis.io/docs/clients/nodejs/
- Confidence: HIGH
## Key APIs
1. `createClient(options)` - Initialize connection
Source: redis.io/docs/clients/nodejs/v4.6#createClient
2. `client.get(key: string): Promise<string | null>` - Retrieve value
Source: redis.io/docs/clients/nodejs/v4.6#get
[...rest of ResearchPack]
✅ Research complete - Ready for planning phase
This skill ensures all implementations are grounded in current, authoritative documentation rather than potentially outdated LLM knowledge.