> For the complete documentation index, see [llms.txt](https://docs.nested-knowledge.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.nested-knowledge.com/get-started/ai-documentation/smart-nest.md). # Smart Nest Learn how to generte a smart nest. Smart Nest is Nested Knowledge's end-to-end review generation workflow. From a single research question, it configures your nest, runs the search, screens abstracts and full texts, applies tags, and runs critical appraisal in one continuous pass, producing a complete, ready-to-review project. Every stage Smart Nest runs is PRISMA-traceable, every AI decision it makes is logged, and every output it produces is editable in **AutoLit** once generation finishes. {% hint style="warning" %} Smart Nest requires an Enterprise pilot or license. See [Start a Pilot Trial](/start-a-pilot-trial/start-a-pilot-trial.md) or set up a [Subscription.](/account-management/subscriptions.md) {% endhint %} *\[Screenshot: Smart Nest generation overview screen]* ### Smart Nest Vs Building A Nest Manually Smart Nest runs the same stages documented as in a Rapid Review, but as one automatic pass instead of a sequence you drive yourself. Choose Smart Nest when you want the fastest path to a defensible first draft and you're comfortable reviewing every stage after generation completes. Choose the manual walkthrough instead when you want to drive a stage yourself — for example, to swap in Robot Screener instead of Smart Screener or to skip a module Smart Nest always runs. See [Conduct a Rapid Review](/get-started/ai-documentation/conduct-a-rapid-review.md). ### Create a Smart Nest {% stepper %} {% step %} ### Name Your Nest On the Nests Homepage, select **Generate Smart Nest** instead of **Create New Nest**. *\[Screenshot: Name your Smart Nest]* {% endstep %} {% step %} ### Add And Refine Your Research Question Enter your research question, then select **Refine** before selecting **Next**. This is the key input for the entire pipeline, so it's worth the extra step: if your question needs refining, a chat box asks you follow-up questions. Answer any or all of them, or select **Save** to continue as written. *\[Screenshot: Refine your research question]* {% endstep %} {% step %} ### Review the generation overview An overview of the Smart Nest generation shows your configuration and lets you check progress at any time. *\[Screenshot: Generation overview]* {% endstep %} {% endstepper %} ### Configuration Smart Nest needs four inputs before it can run. You set the first three in the setup wizard; Smart Nest derives them from your research question unless you edit them: **Research Question** — The refined question that drives the entire pipeline. Smart Nest derives the search strategy, screening criteria, tag hierarchy, and critical appraisal scope from it. **Screening Criteria** — Generated by Smart Config: Screening from your research question, and used for both abstract and full-text screening. **Tag Hierarchy** — Generated by Smart Config: Tagging. Defines what **Adaptive Smart Tags** extracts. **Critical Appraisal System(s)** — Chosen from the systems available in Configure Critical Appraisal, including JBI, SIGN 2011, and SIGN 2019. Review and adjust any of these in the wizard before you start generation. ### What Smart Nest Runs Smart Nest executes the following stages in sequence. Each stage's status and key counts appear in real time on the generation screen. *\[Screenshot: Stage-by-stage progress]* | Stage | What happens | Manual step required? | | ------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------- | | Search | **Smart Search** generates a query from your research question which runs across **PubMed**, **ClinicalTrials.gov** and **OpenAlex,** identifies records, and removes duplicates. | None | | Abstract screening | **Smart Config** generates your screening criteria for abstracts and full texts and **Smart Screener** answers them against each abstract; thresholding advances records that meet every inclusion criterion. | None | | Full-text retrieval | Retrieves available open-access full texts for advanced records automatically. | Yes, after the run. Records without a retrieved full text appear under **Import Remaining** for manual import. | | Full-text screening | **Smart Screener** answers your full-text screening criteria; thresholding includes records that meet every inclusion criterion. | None | | Tagging | **Adaptive Smart Tags** extracts evidence against your configured tag hierarchy for every included study abstract and full text. | None | | Critical appraisal | **Smart Critical Appraisal** generates appraisal answers, with annotated evidence, against your selected system for included studies that have a full text. | None | | Insights | **Smart Insights** generates written summaries with traceable sources for extracted data. | None | | Dashboard/Outputs | Living tables are built in the Dashboard | Yes, table format can be altered to suit desired spreadsheet exports. | ### Review And Edit The Results Once a Smart Nest run finishes, every output behaves like a normal nest: fully editable. The pipeline itself is sequential and committed — you can't reach into a run mid-flight to change a screening decision or a tag. Once it completes, everything is open again. From the generation screen, you can: * **Edit in AutoLit — Show Included Studies** — Opens the included study set in Study Inspector. Re-screen records, edit screening criteria, restructure the tag hierarchy, override extractions, adjust appraisal answers, or re-run any module. Do this before generating Smart Insights or other exports. * **AutoLit Home** — Returns you to the AutoLit homepage and the list of your nests. Once you've reviewed the included studies: * **View in Synthesis — PRISMA** — Opens the PRISMA flow diagram showing record progression through the pipeline. * **View in Synthesis — Smart Insights** — Opens generated Smart Insights, with each claim linked to its source studies and evidence. * **View in Synthesis — Abstract** — Opens the generated abstract and manuscript built from your included studies' data. * **Synthesis Home** — Drops you into Synthesis for downstream analysis and visuals: Qualitative Synthesis, Dashboard, Manuscript, and Critical Appraisal outputs. Because every edit happens after generation, you keep full editorial control over the included set, tags, appraisals, and synthesis outputs. Smart Nest produces the draft; you sign off on the review. See [Navigate Synthesis Outputs.](/review-workflow/synthesis/navigate-synthesis-outputs.md) *\[Screenshot: Generation complete, review actions]* ### Ask Questions With Synthesis Chat A completed Smart Nest pairs naturally with [Synthesis Chat](/review-workflow/synthesis/navigate-synthesis-outputs/synthesis-chat.md), which answers questions using only the data extracted within your nest, with numbered citations back to source studies. Ask it to summarize the evidence on a specific outcome, compare included studies, or draft manuscript text — and because it declines questions your nest's data can't answer, it's well suited to sharing a Smart Nest with colleagues or clients who want to explore the evidence themselves. ### Transparency And Validation Every AI tool in the pipeline above has a published model card, validation evidence and disclosure documentation. See: * [Model cards](/ai-transparency/model-cards/overview.md) * [Disclosure of AI Systems in Nested Knowledge](/policies/policies/artificial-intelligence-policies/disclosure-of-ai-systems-in-nested-knowledge.md) * [Validation of AI Tools](/ai-transparency/ai-in-nested-knowledge/validation-studies.md) ### FAQs
How do I ensure no studies were missed? Smart Nest's search runs across **PubMed**, **ClinicalTrials.gov**, and **OpenAlex** using the strategy generated from your research question, with full PRISMA traceability of records identified, duplicates removed, advanced, and included. To check completeness, review the PRISMA diagram in Synthesis, inspect the search strategy, and consider supplementing with additional searches from proprietary databases. Every screening decision stays reviewable after the run, so you can re-screen any record Smart Screener excluded incorrectly.
What happens if I edit the nest while Smart Nest is running? Smart Nest runs as a sequential, committed pipeline. Editing mid-generation isn't the intended workflow: the generation screen surfaces status only, and editorial actions are blocked or deferred until the run finishes. If you need to change your configuration — research question, criteria, hierarchy, or appraisal system — wait for the run to finish, then edit in AutoLit and re-run the affected module.
Can I change the screening mode or skip certain modules? Not in the wizard. Smart Nest runs the full pipeline as a unit: search, screening, tagging, and critical appraisal all execute in sequence. Module-level selection — skipping tagging, or swapping Smart Screener for Robot Screener — isn't exposed here. For a non-standard pipeline, create a nest manually and run individual modules, or run Smart Nest and then re-run or override the modules you want to handle differently. See [Conduct a Rapid Review.](/get-started/ai-documentation/conduct-a-rapid-review.md)
Will outputs update automatically after I edit included studies? Outputs that read directly from your data — PRISMA diagrams, dashboard visuals, Study Inspector tables, and exports — update automatically when you re-screen, re-tag, or edit appraisals. Smart Insights is different: it's a generated artifact that reflects your data at the moment you generated it, and it won't refresh on its own. To regenerate Insights against an updated included set, return to Dashboard Editor or the Export page and run Smart Insights again. Synthesis Chat reads your nest's current data at the time you ask, so its answers reflect any re-screening, re-tagging, or appraisal edits you've made since generation.
Can I add Meta-Analytical Extraction to a Smart Nest? Not in the current generation pipeline. After Smart Nest completes, configure interventions and data elements in Meta-Analytical Extraction and run Smart MA Extraction manually against your included studies. This keeps quantitative extraction under your explicit configuration, since MA inputs feed forest plots and effect estimates that need deliberate variable definitions.
### Next Steps
:stopwatch:
Conduct a Rapid Review		Run the same stages by hand, module by module.Conduct A Rapid Review
:sliders:
Nest Setup		Choose a review route and prepare each workflow stage.Nest Setup
:user-magnifying-glass:
Start a Pilot Trial		Get access to the full AI tool suite.Start a Pilot Trial
:file-circle-check:
Disclosure of AI		Understand each AI tool's data use, model, and oversight.Disclosure of AI Systems in Nested Knowledge
:scale-balanced:
RAISE Compliance Guide		Plan human oversight and validation for AI-assisted reviews.RAISE Compliance Guide
--- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.nested-knowledge.com/get-started/ai-documentation/smart-nest.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.