# Write a literature review with real citations. From OpenAlex, Crossref and the open web A literature review maps what is established, debated and new on a question, with references a reader can check. With Agentik, your AI finds the key works with OpenAlex, checks recent articles and DOIs with Crossref, searches by meaning with Exa and reads background from Wikipedia. - For: Students, researchers, analysts and writers - Providers: 4 (OpenAlex, Crossref, Exa and Wikipedia), 4 tools - Example (One question, 3 OpenAlex searches, 2 Crossref searches, 2 Exa research searches, 2 Wikipedia summaries): at most $0.05, computed from the live catalog - Updated: 2026-10-10 ## Each provider does one part No single provider covers this job. Through Agentik your AI calls each one for its part, from one connection and one prepaid balance. - **OpenAlex** (1 tool, free): The most cited and most recent scholarly works, with authors and citation counts. - **Crossref** (1 tool, free): DOIs, journals and publication dates for recent articles. - **Exa** (1 tool, at most $0.05): Papers and reports found by meaning, beyond exact keywords. - **Wikipedia** (1 tool, free): Plain-language background on the key concepts. ## Give this to your agent ```text Use my connected Agentik MCP server for this job: literature review. Write a short literature review. First ask me for the research question, the field and the years to cover. Find the most cited and the newest works with openalex/search-works. Check DOIs, journals and recent articles with crossref/search-works. Find research described in other words with exa/search. Get background on the key concepts with wikipedia/page-summary. Finish with a structured review (what is established, what is debated, what is new), a reference list with DOIs, and the open questions. Before running anything, inspect each tool with inspect_tool, ask me for any input you don’t have instead of inventing it, and tell me the total price. Wait for my go. At the end, tell me what Agentik charged. ``` ## You provide - The research question - The field - Years to cover ## What you get: a short literature review with references - What is established, debated and new - A reference list with DOIs - Open questions - Background on key concepts ## How to write a literature review in Claude or ChatGPT Prices are the most each call can reserve for this example: One question, 3 OpenAlex searches, 2 Crossref searches, 2 Exa research searches, 2 Wikipedia summaries. ### 1. Find the key works OpenAlex returns the most cited and the newest scholarly works on your question, with authors and citation counts. - [Scholarly works search](/tools/openalex/search-works.md) `openalex/search-works` (OpenAlex, 3 searches): free ### 2. Check recent articles Crossref checks DOIs, journals and dates for recent articles. - [DOI metadata search](/tools/crossref/search-works.md) `crossref/search-works` (Crossref, 2 searches): free ### 3. Search by meaning Exa searches by meaning for papers described in other words. - [Exa web search](/tools/exa/search.md) `exa/search` (Exa, 2 searches): at most $0.05 ### 4. Get the background Wikipedia gives plain-language background on the key concepts. - [Wikipedia page summary](/tools/wikipedia/page-summary.md) `wikipedia/page-summary` (Wikipedia, 2 concepts): free ### 5. Write the review Your AI writes the review with a reference list and the open questions. Example total: at most $0.05. Each call is billed for measured usage; unused reservations are released. ## Ask it your way - What does the research say about sleep and memory consolidation? - Find the ten most cited papers on remote work productivity since 2020. - Write a short literature review on microplastics in drinking water with DOIs. - What is still debated about intermittent fasting? ## Why through Agentik - Scholarly databases, web search and Wikipedia come through one connection. - OpenAlex, Crossref and Wikipedia are free; the example’s cost is two Exa searches. ## Good to know - OpenAlex, Crossref and Wikipedia are free; the example’s cost is Exa’s searches. - Your AI reads abstracts and metadata; full texts behind paywalls are not read. ## When to use something else For a systematic review with screening and bias assessment, use a reference manager and review software. This playbook gives a sourced first map of a field. ## Literature review: questions ### Does it read full papers? It reads abstracts and metadata; full texts behind paywalls are not read. ### Can I use it for a systematic review? As a first map. A systematic review needs screening and bias assessment in dedicated software. ### What does the example cost? At most $0.05, reserved before each call and settled at the measured usage. ## Keep going - [Fact-check a claim](/fact-check.md): A claim checked against the web, the news, research papers and Wikipedia, with the original source found in the Internet Archive. - [Expert finder](/expert-finder.md): Researchers and practitioners who published or spoke recently on your topic, their papers and podcast appearances, their X profiles and emails. - [Learning plan](/learning-plan.md): Any topic broken into what to learn first, with plain-language background, the key papers, the books and the best lectures, in a week-by-week plan. - [Research tools for your agent](/tools/research.md) - [Knowledge tools for your agent](/tools/knowledge.md) - [Web search tools for your agent](/tools/search.md) [All playbooks](/playbooks.md) · [Connection guides](/docs/clients.md) · [Billing](/docs/billing.md) --- [HTML page](/literature-review) · [Agent index](/llms.txt)