External Sources - Research Staging Area

Purpose

This directory serves as a temporary staging area for external research and documentation that hasn't yet been incorporated into the structured documentation framework.

Think of this as a "research inbox" where you can quickly drop materials from other platforms, allowing Claude Code and other LLMs to access this context during feature planning sessions.

How It Works

1. Drop Research Here

When researching a new feature or technology, you might gather information from:

• Blog posts and articles
• Vendor documentation excerpts
• API documentation snippets
• Competitor feature analyses
• Academic papers or methodologies
• Stack Overflow discussions
• GitHub issues and discussions
• Conference talks or presentations
• Internal notes from other platforms

Just drop these materials here as markdown files without worrying about structure or formatting.

2. LLM Context Access

Claude Code can now:

• Read all materials in this directory
• Ask clarifying questions about the research
• Help identify which parts are relevant
• Suggest how to incorporate findings into structured docs
• Generate Q&A sessions based on the research

3. Progressive Refinement

During feature planning, Claude can help you:

1. Review the raw research materials
2. Extract relevant information
3. Ask questions to clarify understanding
4. Incorporate findings into:
5. Feature Briefs
6. Technical Plans
7. ADRs
8. External reference links

4. Cleanup

Once research is incorporated into structured documentation:

• Delete the source files (they've served their purpose)
• Or move to an archive folder if you want to keep them
• The structured docs become the source of truth

File Naming Convention

Use descriptive names that help identify the content and date:

YYYY-MM-DD-{topic}-{source}.md

Examples:
2025-11-21-oauth2-auth0-docs.md
2025-11-21-payment-stripe-api-notes.md
2025-11-21-systematic-review-prisma-guidelines.md
2025-11-21-competitor-covidence-features.md

Example Workflow

Step 1: Research Collection

# Quick Research Notes - Payment Processing

## From Stripe Docs
[Paste relevant sections]

## From Auth0 Integration Guide
[Paste relevant sections]

## Stack Overflow Discussion
[Paste helpful answers]

## Questions/Thoughts
- How does this apply to our use case?
- What are the security implications?

Step 2: LLM-Assisted Analysis

User: "Claude, review the payment processing research in /docs/sources/ and help me understand the key decisions we need to make"

Claude: *Reads all source files, asks clarifying questions, identifies decisions*

Step 3: Structured Documentation

Claude helps create:

• Feature Brief with clear requirements
• ADR for payment provider selection
• Technical Plan with implementation details

Step 4: Cleanup

Delete source files after incorporation into structured docs.

What This Is NOT

This is not a permanent reference library. It's specifically:

• ❌ NOT for long-term storage of external docs
• ❌ NOT for organized API references
• ❌ NOT for official technical standards
• ❌ NOT a replacement for proper documentation

This is:

• ✅ A temporary staging area
• ✅ An LLM context provider
• ✅ A research inbox
• ✅ A place for unprocessed materials

Best Practices

1. Keep it temporary - Delete files after incorporating into structured docs
2. Don't organize - Just drop files here for LLM consumption
3. Use descriptive names - Include date and topic for easy identification
4. Regular cleanup - Review weekly and delete processed materials
5. Focus on relevance - Only include materials relevant to current work

Lifecycle

graph LR
    A[External Research] -->|Copy/Paste| B[/docs/sources/]
    B -->|LLM Reads| C[Q&A Sessions]
    C -->|Extract & Refine| D[Structured Docs]
    D -->|Cleanup| E[Delete Sources]

Current Contents

This directory should usually be empty or contain only materials for active feature planning.

File	Date Added	Purpose	Status
(empty)	-	-	-

---

Remember: This is a staging area, not a filing cabinet. Process and delete regularly.