# Researcher > Researcher Skill - Skill: `adaptivex-gh/researcher` (Agent Skill) - Install (CLI): `npx skillmds@latest add adaptivex-gh/researcher` - Raw SKILL.md: https://api.skillmd.com/api/skills/adaptivex-gh/researcher/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: adaptiveX-gh (https://skillmd.com/u/adaptivex-gh) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/adaptivex-gh/researcher --- # Researcher Skill Guide users through researching identified knowledge gaps in course plans using the CRAAP test and 4-tier source hierarchy. ## When to Use After running `coursekit.plan`, if research needs are identified, this skill guides the user through: 1. Understanding the research needs 2. Finding credible sources 3. Evaluating source quality 4. Documenting findings 5. Updating the course plan with research ## Prerequisites - Course plan has been created with `coursekit.plan` - Research needs have been identified (stored in `.coursekit/research_needs.json`) - User is ready to conduct research ## Stage 1: Review Research Needs First, call the `coursekit.research` MCP tool to understand what research is needed: ``` Use coursekit.research to review all identified research needs ``` Then present the research needs to the user, grouped by: - **Priority**: Critical > High > Medium > Low - **Type**: Data, Example, Evidence, Concept Ask the user: > I've found [N] research needs for your course plan. The highest priority items are: > > [List top 3-5 critical/high priority needs] > > Would you like to: > 1. Research these systematically (recommended) > 2. Focus on critical/high priority only > 3. Skip research and proceed with current information ## Stage 2: Research Each Need For each research need, guide the user through this process: ### 2.1 Understand the Need Present the specific research need: ``` Need: [need.need] Type: [need.type] Priority: [need.priority] Context: [need.context] Suggested search: [need.suggested_search] ``` ### 2.2 Source Quality Framework Explain the 4-tier source hierarchy for this type of need: **For Data/Evidence needs:** - **Tier 1 (Best)**: Peer-reviewed academic papers, meta-analyses, systematic reviews - **Tier 2 (Good)**: Industry reports from reputable firms, government statistics, academic books - **Tier 3 (Fair)**: Practitioner blogs from recognized experts, conference talks, reputable tech blogs - **Tier 4 (Verify)**: General blogs, social media, marketing materials (use with caution) **For Example/Concept needs:** - **Tier 1 (Best)**: Official documentation, academic textbooks, established tutorials - **Tier 2 (Good)**: GitHub repos with high stars, case studies from known companies - **Tier 3 (Fair)**: Blog posts from practitioners, Stack Overflow high-voted answers - **Tier 4 (Verify)**: Random tutorials, unverified examples (validate thoroughly) ### 2.3 CRAAP Test For each source the user finds, apply the CRAAP test: **Currency**: When was it published? Is it current for this topic? - Data/stats: Prefer last 2 years - Examples: Prefer last 3 years unless classic - Concepts: Foundational knowledge can be older **Relevance**: Does it directly address the need? - Exact match or requires interpretation? - Appropriate depth for target audience? **Authority**: Who created it? What are their credentials? - Author's expertise in the field - Institutional affiliation - Publication venue quality **Accuracy**: Is the information reliable and verified? - Claims supported by evidence - References to other credible sources - Transparent about limitations **Purpose**: Why was it created? Any bias? - Educational vs. marketing intent - Potential conflicts of interest - Balanced perspective ### 2.4 Document Findings For each validated source, collect: ```json { "research_id": "research_X", "source_url": "https://...", "source_title": "Title of source", "source_type": "academic paper|industry report|blog|documentation|etc", "tier": 1-4, "craap_score": { "currency": "high|medium|low", "relevance": "high|medium|low", "authority": "high|medium|low", "accuracy": "high|medium|low", "purpose": "high|medium|low" }, "key_findings": [ "Specific finding 1", "Specific finding 2" ], "quotes": [ "Direct quote if useful" ], "how_to_use": "How this addresses the research need" } ``` ## Stage 3: Save Research Notes After researching each need, update the research notes file: Create or update `.coursekit/research_notes.json`: ```json { "findings": [ { "research_id": "research_1", "need_id": "research_1", "sources": [ { "url": "https://example.com/paper", "title": "Study on X", "type": "academic paper", "tier": 1, "craap_score": { "currency": "high", "relevance": "high", "authority": "high", "accuracy": "high", "purpose": "high" }, "key_findings": ["Finding 1", "Finding 2"], "quotes": ["Quote if relevant"], "accessed_date": "2025-11-09" } ], "synthesis": "Summary of what was learned and how it addresses the need", "recommendation": "How to incorporate this into the course plan", "status": "completed" } ], "metadata": { "totalFindings": 1, "researchedBy": "User name", "lastUpdated": "2025-11-09T00:00:00.000Z" } } ``` After documenting findings, also update `.coursekit/research_needs.json` to mark the need as completed: ```json { "id": "research_1", "type": "data", "priority": "high", "need": "...", "context": "...", "suggested_search": "...", "location": "...", "status": "completed" // Changed from "pending" } ``` ## Stage 4: Integrate Findings into Plan After research is complete, guide the user to: 1. Review all findings 2. Identify how to integrate them into the plan: - Add statistics to strengthen claims - Include examples in materials list - Reference evidence in rationale - Expand on concepts with sources 3. Suggest specific plan updates: ``` Based on your research, I recommend updating the plan: Module 2: - Add citation: [Author, Year] found that [statistic] - Include example: [Company] case study (source: [URL]) - Materials: Add reference to [documentation] Module 3: - Strengthen claim about [X] with evidence from [source] ``` 4. Ask if user wants to regenerate the plan with research incorporated, or manually update it ## Stage 5: Summary Provide a research summary: ``` Research Summary: - Total needs identified: [N] - Needs researched: [N] - Sources found: [N] - Tier breakdown: - Tier 1 (peer-reviewed/official): [N] - Tier 2 (reputable industry): [N] - Tier 3 (practitioner): [N] - Tier 4 (verify carefully): [N] All findings saved to .coursekit/research_notes.json Next steps: 1. Review research_notes.json for complete findings 2. Update course plan to incorporate research 3. Run coursekit.plan again to regenerate with improvements (optional) 4. Proceed to coursekit.tasks to begin content development ``` ## Best Practices 1. **Prioritize**: Start with critical and high priority needs 2. **Quality over Quantity**: One Tier 1 source beats five Tier 4 sources 3. **Document thoroughly**: Future you will thank you for good notes 4. **Be skeptical**: Always apply CRAAP test, even to sources you like 5. **Synthesize**: Don't just collect links, extract actionable insights 6. **Cite properly**: Note URLs, authors, dates for proper attribution 7. **Know when to stop**: Diminishing returns after 3-5 good sources per need ## Error Handling If `coursekit.research` returns "No research needs found": - Verify that `coursekit.plan` has been run - Check if `.coursekit/research_needs.json` exists - The plan might not have identified any gaps (rare but possible) If user can't find good sources: - Suggest broadening search terms - Consider adjacent topics or foundational concepts - Check if the need is actually critical or can be addressed with general knowledge - Recommend consulting with subject matter experts ## Output Files This skill helps create/update: - `.coursekit/research_notes.json` - All research findings with sources - `.coursekit/research_needs.json` - Updated to mark completed research These files are then used by subsequent phases (tasks, implement) to ensure course content is well-researched and credible.