Multi-Step Research and Summarize Workflow
The workhorse of the category. Plans queries, executes across sources, cross-references, drafts structured brief with inline citations. Configurable depth (brief, standard, deep).
Every AI Skill in our directory that produces sourced research — academic paper summaries, market briefs, competitive intelligence, news scans, legal research memos. Each Skill enforces citation discipline (nothing invented, everything traceable) and has been tested against real research questions. Provider-agnostic — same Skill runs on Claude, GPT-5, Gemini, Grok, or Cohere.
Research & Analysis Skills accept a research question and produce a sourced answer. Under the hood they orchestrate multi-round work: plan queries, execute against web sources or specific databases, cross-reference findings, and produce a structured brief with inline citations. The best of these Skills pair with a search MCP server (Perplexity, Brave, or Grok's DeepSearch) for the fetch phase and use the AI provider for the synthesis phase.
The value the Skill layer adds is citation discipline — every non-obvious claim in the output traces to a source URL. Skills refuse to invent citations, mark single-source claims as needing verification, and surface contradictions between sources rather than silently averaging them. For research that will be quoted, cited, or acted on, this discipline is the difference between AI drafts you can use and AI drafts that get you in trouble.
What lives here vs. elsewhere: if the output is a sourced research artifact (a brief, a summary, a memo), it belongs here. If it's a document with formatting that contains research (a formal report, a case-study PDF), see Document Creation. If it's downstream content built from research (a blog post, an article), see Content Creation. If it's business-domain research (sales prospecting, HR intel), see Business Operations.
Our editorial picks — heaviest tested, highest reader ratings.
The workhorse of the category. Plans queries, executes across sources, cross-references, drafts structured brief with inline citations. Configurable depth (brief, standard, deep).
Turn dense papers into structured briefs — abstract, methods, findings, limitations, implications. Preserves figure and citation refs. Batch mode for literature reviews.
Generate a sourced company profile for sales, investment, or diligence. Aggregates from public sources with cited claims. Refuses to fabricate.
Filter by sub-type, search by name, sort by recency or alphabetical.
Autonomous multi-round research. Plans queries, executes via web search MCP, cross-references, drafts structured brief with citations.
Given a corpus of papers or reports on the same intervention, extract effect sizes into a comparison table. Flags outliers.
Given a long brief or report, produce a tight 250-word executive summary preserving the most decision-relevant claims.
Summarize papers into structured briefs — abstract, methods, findings, limitations, implications. Preserves figure and citation refs.
Given 20-100 paper summaries, produce a cross-corpus synthesis: themes, contradictions, gaps, follow-up literature.
Given a seed paper, map its citation network and rank related papers by relevance to your research question.
For empirical papers, check whether methods are described in enough detail to reproduce. Flags missing seeds, hyperparameters, data.
Given a product category, estimate TAM/SAM/SOM using bottom-up and top-down methods. Cites source assumptions.
Given a category, identify direct and adjacent competitors. Positioning matrix, feature comparison, funding snapshot.
Analyze trends in an industry over a period. Combines research reports, news, filings. Distinguishes signal from noise.
Summarize 10-K, 10-Q, 8-K, S-1 filings. Highlights risk factors, MD&A, material events. Comparison across periods.
Analyze a news article for bias, factual claims, source quality. Distinguishes reporting from opinion.
Configurable daily digest for a specific topic or beat. Deduplicates, ranks by relevance, cites sources.
Cluster news articles about the same event across outlets. Surfaces framing differences and factual disagreements.
Cross-check claims in a press release against independent sources. Flags unverified or overreaching claims.
Search case law for a specific legal question. Returns relevant cases with holdings and quoted passages. Cites Westlaw/CourtListener.
Monitor regulatory changes (SEC, FTC, EU, GDPR) affecting a specific business area. Weekly digest with source links.
Answer a compliance question across specified jurisdictions. Flags where jurisdictions differ. Not a substitute for counsel.
MCP servers our team pairs with Research & Analysis Skills for full workflows.
Recommended search backend for citation-first research workflows.
Cost-conscious search backend. Great for high-volume research jobs.
Save research briefs directly to a Notion research workspace.
Save briefs to a shared research folder in Drive.
Post daily digests to team channels or notify on specific triggers.
Store research briefs in enterprise Confluence spaces.
Longer-form tutorials for the concepts behind these Skills.
How to choose the right depth budget for a research question — trade-offs of cost, time, and quality.
Head-to-head comparison of the two leading AI-native research modes.
How to restrict research to primary sources when reliability matters more than coverage.
Head-to-head reviews that touch on Research & Analysis workflows.
Head-to-head across 30 research questions. Which produces better-cited output?
Which provider synthesizes multi-source research most reliably?
Skill-based vs provider-native research modes: when each wins.
Our sister sites cover errors and code recipes across the AI ecosystem. Bookmark them for when a Skill hits an edge case.
Weekly roundup of new Skills we've tested, install notes, and workflow patterns. Join 2,400+ readers of "The Skill Report" every Tuesday.