Research and Learning skills
Research and Learning Skills-md for Claude, Codex & AI Agents
Research and learning skills are SKILL.md files that turn Claude, Codex and other AI agents into a careful research partner and learning coach. They frame research questions, search and evaluate sources, fact-check claims, use AI research tools safely, run interviews and surveys, synthesise findings into insights and build learning plans with practice, spaced repetition and explanation.
13 of 13 skills published, version 1.0.0, October 2026.
Research Planning
Plan research before searching: the decision it serves, a sharp bounded question, an issue tree of sub-questions, hypotheses, depth matched to stakes, methods from desk research to interviews, surveys and experiments, sources, evidence standards, confidence labels, time plans and agreed outputs.
Desk Research
Run effective desk research: search terms from sub-questions, source types from official statistics to filings and studies, search operators, primary sources, skimming and reading, lateral credibility checks, a research log, tracing citations, triangulation and summaries of what is known, uncertain and unknown.
Source Evaluation
Evaluate sources before relying on them: lateral reading and the SIFT method, deeper checks of authority, method, evidence, currency, purpose and transparency, red flags for statistics, surveys, studies, reports, news, websites, social media and AI outputs, and confidence labels with reasons.
Fact-Checking
Fact-check documents, decks, articles and AI outputs: extracting every claim, prioritising by risk, verifying against primary sources, recalculating numbers and units, confirming names, dates and quotes, spotting misleading-but-true framing, fixing wording and keeping a fact-check log.
AI-Assisted Research
Use AI research tools well: choosing between deep research modes, AI search, assistants and browsing agents, writing research briefs with preferred sources and evidence standards, verifying citations, filling gaps, resolving conflicts, protecting confidential data and writing a human-checked synthesis.
Literature Review
Review research evidence for work and study questions: question framing and inclusion criteria, review types from quick scans to systematic reviews, database searching and search strings, screening, data extraction, quality assessment, synthesis by theme and write-ups with confidence and gaps.
Expert and User Interviews
Plan and run customer, user, buyer, churn and expert interviews: learning goals, recruiting by behaviour, discussion guides with past-behaviour questions, good versus leading questions, consent and recording, note-taking, debriefs, expert interview tips and handover to synthesis.
Survey Design
Design surveys that produce trustworthy data: objectives tied to decisions, population and sample size with margins of error, neutral single-idea questions, answer formats and scales, ordering and skip logic, pilots, invitations and reminders, cleaning, segment analysis and reports with limitations.
Research Synthesis
Turn research into insight: organising and coding observations, affinity mapping, insight statements, counting and weighing evidence, tensions and surprises, frameworks such as journeys and jobs, implications and prioritised recommendations, and presentations that answer the research question first.
Active Reading and Notes
Read actively and take useful notes: a purpose and questions, previewing, reading depths from skim to study, questioning the text, recall instead of rereading, notes in your own words, accurate quotes and sources, summaries, linking notes to projects and reviewing them.
Learning Plans
Build learning plans that work: testable target performances, starting assessments, sub-skills, focused resources, deliberate practice with feedback, real projects, weekly schedules, retrieval practice, spacing and interleaving, progress checks and adjustments, with a 12-week example.
Spaced Repetition and Retrieval Practice
Remember what you learn with retrieval practice and spaced repetition: choosing what is worth remembering, understanding before memorising, one-fact flashcards and cloze deletions, Anki or Leitner systems, daily review routines, practice questions, card maintenance and application at work.
Explain to Learn
Learn by explaining: self-explanation research and the Feynman technique, explaining without notes, finding and studying gaps, plain-language rewrites, examples and non-examples, analogies with limits, teaching others, explanation templates and using AI as a questioning partner.
No skill matches that search. Clear it or choose All.
What these skills do
AI makes it easy to produce research that looks finished: long reports, dozens of citations and confident conclusions. Without method, those reports answer vague questions, rely on weak sources and include claims nobody checked. Learning has a similar trap: watching courses and highlighting books feels productive but fades quickly.
These skills bring method to both. Research starts with a decision and a sharp question, uses primary sources, checks credibility by reading laterally, verifies every important claim and ends with insights tied to evidence and confidence levels. Learning focuses on testable goals, practice with feedback, retrieval and spaced review and explaining ideas simply, the techniques learning research rates most effective.
Research methods and learning techniques
The set covers planning, desk research, source evaluation, fact-checking, AI research tools, literature reviews, interviews, surveys, synthesis, active reading, learning plans, spaced repetition and learning by explaining.
Every skill keeps you in charge: the agent plans, searches, drafts and quizzes, while you decide what counts as enough evidence and do the practice that turns knowledge into skill.
Where to start: a question or a skill
If you need an answer, start with research-planning. Write the decision the research serves, a sharp question and three to seven sub-questions, then choose methods: desk-research for what already exists, expert-user-interviews and survey-design for what only people know, and literature-review for what studies show. Check sources with source-evaluation, verify claims with fact-checking and finish with research-synthesis.
If you need to learn, start with learning-plans. Define a testable target, spend most of your time practising with feedback, add spaced-repetition for knowledge you need to recall quickly and use explain-to-learn to check you really understand. active-reading-notes turns books and reports into notes you can use.
AI tools speed up both. ai-assisted-research shows how to brief deep research tools and verify what they return, so speed does not come at the cost of accuracy.
A typical research project using these skills runs like this: an hour to plan the question and sub-questions, a day of desk research and AI-assisted searching with every claim logged, a few expert or customer interviews to fill gaps that published sources cannot, a synthesis session to turn findings into insights with confidence labels, and a final fact-check before the summary goes to decision makers. A typical learning project runs for six to twelve weeks with short daily practice, weekly self-tests, spaced reviews of key knowledge and a real project that proves the new skill. Both follow the same idea: clear goals, reliable evidence or practice, and honest checks of what is known and what is not.
The 13 skills and the research behind them
All thirteen skills in this set are live.
| Skill | What it does | Status |
|---|---|---|
research-planning | Frame the question, scope and methods before searching | Live |
desk-research | Search precisely and find primary sources | Live |
source-evaluation | Judge credibility with lateral reading and SIFT | Live |
fact-checking | Check every claim before it goes out | Live |
ai-assisted-research | Brief AI research tools and verify what they return | Live |
literature-review | Find and synthesise what research says | Live |
expert-user-interviews | Interviews that uncover real behaviour | Live |
survey-design | Surveys that produce trustworthy numbers | Live |
research-synthesis | Turn findings into insights and recommendations | Live |
active-reading-notes | Read with purpose and keep useful notes | Live |
learning-plans | Learn skills through practice and feedback | Live |
spaced-repetition | Remember what you learn with retrieval and spacing | Live |
explain-to-learn | Understand deeply by explaining simply | Live |
The research behind these skills
| Source | How the skills use it |
|---|---|
| Lateral reading (Wineburg and McGrew) and SIFT (Caulfield) | Check sources by reading what others say about them |
| Dunlosky and colleagues, learning techniques review | Practice testing and spacing over highlighting and rereading |
| Roediger and Karpicke; Cepeda and colleagues | Retrieval practice and spaced review for long-term memory |
| Ericsson on deliberate practice | Focused practice on weaknesses with feedback |
| Chi on self-explanation | Explaining to understand |
| The Mom Test (Fitzpatrick) | Interview questions about past behaviour |
Example prompts
Each skill triggers on everyday wording. These prompts are good starting points:
- Plan: “Plan research on whether we should enter Kenya.”
- Sources: “Is this statistic trustworthy?”
- Fact-check: “Fact-check this investor update.”
- AI research: “Write a brief for a deep research tool on African payments.”
- Interviews: “Write an interview guide for churned customers.”
- Learning: “Build a 12-week plan to learn SQL for marketing.”
- Memory: “Turn my notes into Anki flashcards.”
What the agent needs from you
Good research and learning help depends on a few inputs:
- The decision or goal: what the research informs or what you want to be able to do.
- Scope and time: depth, deadline and hours per week.
- Access: sources, databases, customers, experts or courses.
- Standards: how much evidence is enough and how confident you need to be.
- Your starting point: what you already know.
If something is missing, the skills ask only what changes the plan.
Sharing these details at the start saves rounds of back-and-forth and keeps the work focused on what matters.
What an agent cannot do, and myths to ignore
An agent can search, summarise, organise and quiz, but it cannot guarantee that sources are accurate or that citations say what is claimed, and it cannot do the practice that builds skills. Verification and practice stay with you.
| Myth | What works better |
|---|---|
| More sources mean better research | Fewer, stronger sources that answer the question |
| Citations prove a claim | Open the source and check it says what is claimed |
| Highlighting and rereading help you learn | Self-testing and spaced review work far better |
| Watching courses builds skills | Practice with feedback on real tasks |
| Customers can tell you what they will buy | Ask about what they did, not what they might do |
Frequently asked questions
How do I plan research?
Start with the decision, write a sharp question, break it into sub-questions and choose methods and sources for each.
How do I know if a source is reliable?
Read laterally with the SIFT method, check method, date and purpose and trace claims to the original.
Can I trust AI research tools?
Use them to find and draft, then verify key claims by opening the sources.
What is the best way to learn a new skill?
Define a testable goal and practise with feedback, adding retrieval and spaced review.
Does spaced repetition really work?
Yes, research on spacing and retrieval practice shows strong benefits for long-term memory.
What is the Feynman technique?
Explaining a concept simply, finding gaps, studying them and explaining again.
Which agents can use these skills?
Claude, Claude Code, Codex, Cursor, Gemini CLI and other tools that support the Agent Skills format.
Are these skills free?
Yes, under CC BY 4.0, as long as you credit Afeez Adebayo and link to the source page.
Sources and more skills
- Dunlosky et al. (2013), Improving students' learning with effective learning techniques
- Roediger and Karpicke (2006), test-enhanced learning
- Wineburg and McGrew, lateral reading research
- Mike Caulfield, SIFT method
- Rob Fitzpatrick, The Mom Test
Checked on 14 October 2026. See all Productivity skills, including AI Workflows and Automation.
