> 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/verifying-entries.md). # Verifying entries ## Two checkpoints, before and after entry Filed gives you a chance to catch problems at two different points: before the data-entry bot ever touches tax software, and after it finishes. Both matter, and they catch different kinds of issues. ## Before data entry: the Forms tab Before the bot runs, the [Forms tab](/reference/products/ai-tax-prep/forms-tab.md) shows every form and screen the AI plans to prepare, along with the source document behind each one. This is the moment to review the AI's plan - open a row's source reference, confirm it's reading the right document, and exclude anything that shouldn't be part of the return - before that plan gets written into tax software. See [Reading the form list](/reference/products/ai-tax-prep/forms-tab.md#reading-the-form-list) and [Excluding a form](/reference/products/ai-tax-prep/forms-tab.md#excluding-a-form). [Pre-Entry Notes](/reference/products/ai-tax-prep/pre-entry-notes-tab.md) is the companion review surface at this same stage: conflicts, missing documents, and anything the AI set aside, all surfaced before entry fires. ## How do I know everything exported correctly? After the data-entry bot fills a return, Filed does not just assume the values landed correctly. It reads each field back out of the tax software and compares it against what it meant to enter. This is the verification pass referenced as "Verifying" in the pipeline view (see [Workflow overview](/reference/products/ai-tax-prep/workflow-overview.md#automated-data-entry)), and its result is what drives the Confirmed / Flagged / Pending / Manual counts on the [Summary tab](/reference/products/ai-tax-prep/summary-tab.md). ## What gets compared For every field the bot enters, Filed tracks two things: * **The intended value** - what Filed extracted from the source documents and meant to write into that field. * **The verified value** - what Filed reads back from the tax software field after entry. If the two match, the field is marked Confirmed. If they differ, the field is Flagged for review. Values are compared with normal tax-value normalization, so formatting differences alone (commas, cent rounding, currency symbols) do not cause a false mismatch. ## Where to review a flagged field A flagged field does not mean the AI entered the wrong number; it means the value written into tax software did not read back the way Filed expected, and a human should confirm which value is correct. Open the flagged form from the Summary tab's forms list or from the Forms tab directly, and the Binder shows the AI's source citation for the field alongside the value that was entered. Compare the source document to the entered value to decide whether to correct it in tax software or clear the flag. See [Forms tab](/reference/products/ai-tax-prep/forms-tab.md) for how source references work on each row. ## Why a field might be Pending instead of Confirmed or Flagged Pending means the field is in scope for verification but the pass has not resolved it yet. This can happen while the bot is still running, or for fields where the tax software's own recalculation (for example, a computed subtotal) has not settled at the time Filed reads the value back. Pending fields are not errors; revisit the Summary tab after the pipeline finishes to see whether they resolved to Confirmed or Flagged. ## Manual fields are not verified Manual fields come from form types the data-entry bot does not currently support, so nothing was entered for Filed to verify. These must be entered directly in tax software; see the "Manual" row on the [Summary tab](/reference/products/ai-tax-prep/summary-tab.md#after-data-entry-completes). ## Relationship to backups Verification confirms individual field values; it does not undo an entry. If a flagged or incorrect value needs to be rolled back, use the Backups panel described in [Summary tab](/reference/products/ai-tax-prep/summary-tab.md#the-backups-panel) to restore the return to a prior state rather than correcting fields one at a time. --- # 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/verifying-entries.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.