> For the complete documentation index, see [llms.txt](https://help.filed.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.filed.com/reference/products/ai-tax-prep/summary-tab.md). # Summary tab The Summary tab reflects the client's current position in the tax prep pipeline. What it displays changes as the pipeline moves through its stages: processing, ready for data entry, and complete. ## While processing While the pipeline is running, the page shows a live progress view. Each pipeline phase (binder creation, extraction, reconciliation, planning) appears as a step with a spinner while active and a check mark when complete. The preparer can navigate away; the view updates automatically when the next phase begins and when the pipeline finishes. ## After extraction: ready for data entry When extraction and planning are complete but the data-entry bot has not run yet, the Summary shows the current state as a task result. Three counts are shown: * **Forms** - the number of tax screens the AI built a data-entry plan for. * **Documents** - the number of uploaded source documents the AI extracted from. * **Gaps** - fields or screens the AI could not fully resolve, also visible in Pre-Entry Notes. The Backups panel is present from this point forward. When a prior-year return was imported, an "Initial" backup appears here showing the state of the tax software return before Filed ran anything. This is the correct moment to review the [Forms Tab](/reference/products/ai-tax-prep/forms-tab.md) and [Pre-Entry Notes Tab](/reference/products/ai-tax-prep/pre-entry-notes-tab.md) before triggering data entry. ## After data entry completes When the data-entry bot finishes, the Summary shows a completion view with an animated check mark and a "Tax prep complete" badge. The view reports how many forms were entered and how many fields were confirmed, then presents four stat cards: | Card | What it counts | | --------- | ---------------------------------------------------------------------------------------------------------- | | Confirmed | Fields the bot entered and the post-entry validation pass confirmed as correct. | | Flagged | Fields the bot entered but that the validation pass flagged for human review. | | Pending | Fields that are in scope but were not yet confirmed by the validation pass. | | Manual | Fields from form types the bot does not currently support; these must be entered manually in tax software. | When Flagged is greater than zero, an amber alert banner appears: "N fields flagged for review - Open the forms in the binder to review flagged entries." ### Draft return PDF When the tax software produces a draft return PDF after entry, the Summary embeds a preview panel showing the PDF. Two buttons appear above it: * **Run AI Review** - starts an AI review pass on the draft return to check for issues. * **Review results in binder** - opens the Binder's Forms screen to review entered values against source documents. ### Forms list Below the stat cards, the completed view lists each entered form with its confirmed-field count (green check mark) and flagged-field count (amber alert icon). Clicking any row opens the form in the Binder. ## The backups panel The Backups panel, accessible via the archive icon in the Summary, shows automatic snapshots of the tax software return. Three backups are created during a full AI Tax Prep run: | Backup label | When it is created | | ----------------- | ------------------------------------------------------------------------------------------------ | | Initial | Before Filed runs anything - the state of the return as it existed in tax software at the start. | | Before data entry | Immediately before the bot begins filling fields. | | After data entry | After the bot finishes. | Each backup has a download button. For desktop and file-based integrations, the download can be a native tax software backup. ProConnect produces a structured recovery snapshot instead; it cannot be imported through the ProConnect user interface. The most important recovery scenario: if a problem is discovered after the bot runs, use the **Before data entry** backup to identify the pre-Filed state. For ProConnect, contact Filed support for an exact-run restore; see [ProConnect backups and recovery](/reference/plugins/integrations/tax-software/proconnect-tax/proconnect-backup-and-recovery.md). The Backups panel is visible whenever a backup exists - while the bot is running, while the pipeline is idle between phases, and after the pipeline completes. ## Admin-only controls A gear icon and a "View Playbook" button appear in the Summary for admin-level users. The gear opens firm-level preferences for AI Tax Prep. "View Playbook" shows the firm's rules for automatically routing forms to tax software. These controls are not shown to preparers at lower permission levels. ## What to do after reviewing the Summary tab * **If Flagged is greater than zero:** click "Review results in binder" and inspect each flagged field against its source document. The Binder shows the AI's citation for every entered value. * **If Manual is greater than zero:** open tax software and enter the listed form types manually. The Summary lists which forms are unsupported. * **When a draft return is available:** click "Run AI Review" to run an AI check on the completed return. * **If something went wrong during entry:** open the **Before data entry** backup and follow the restore method for the connected tax software. ProConnect restores are support-assisted. ## TaxPrepSummaryStatus values The following status values drive what the Summary displays. They are exposed in the GraphQL API and may be useful for integrations or support troubleshooting: | Status | Meaning | | ------------------ | ------------------------------------------------------------------------------------------------------------------------------- | | READY\_TO\_PLAN | Extraction complete; reconciliation has not run yet. | | READY\_TO\_ENTER | Reconciliation complete; the data-entry bot has not run yet. | | NOTHING\_TO\_ENTER | Reconciliation ran but produced no actionable per-field targets. | | PLAN\_UNSUPPORTED | A data-entry plan was built, but every planned form is outside the bot's supported list. Manual entry is required. | | NEEDS\_REVIEW | The bot ran and produced results, but some entries need human review. | | PARTIAL\_MANUAL | The bot completed everything it could support, but some planned forms are outside its support list and still need manual entry. | | READY | The bot ran successfully and all entries are confirmed. | --- # 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 dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://help.filed.com/reference/products/ai-tax-prep/summary-tab.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. 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.