Where to See Your AI Review History in zipBoard
Open your PDF in the markup viewer and select the AI Review tab. Orbit Check retains a complete, persistent audit trail of every review session. You can see both Accepted findings (converted into active zipBoard tasks and comments) and Rejected findings (dismissed suggestions) at any time, complete with their exact visual locations on the document.
Without a clear audit trail, teams end up re-reviewing the same sections, second-guessing past decisions, or losing sight of dismissed suggestions. As an AI review agent, Orbit Check maintains complete review transparency. Orbit Check retains your full review history directly within the document workspace, keeping both Accepted and Rejected findings permanently recorded so your team always has access to a verifiable audit log.
Orbit Check is in beta on paid plans and free to use until 15 January 2027. We’ll announce pricing in advance, and you won’t be charged without choosing to.
Why Auditability Matters in AI Document Reviews
Maintaining a persistent history of AI findings provides three major operational benefits:
- Eliminates Duplicate Work: Team members can quickly check whether an issue was previously evaluated and rejected before re-flagging it.
- Full Compliance & Accountability: Enterprise and regulated workflows (such as architectural plan sets or technical user manuals) require a clear record of who validated or dismissed each AI recommendation.
- Context Preservation: Even if a suggestion is rejected, the historical record retains Orbit Check's original reasoning and the precise visual coordinates on the PDF page.
How to Access and Inspect Your AI Review History
Follow these steps to navigate past Orbit Check reviews and filter findings by status inside the zipBoard markup viewer:
- Navigate to Your PDF Asset
In your zipBoard project, open the PDF document that underwent an AI review.
- Open the AI Review Tab

In the right-hand panel of the markup viewer, click on the AI Review tab. This tab serves as the central hub for all Orbit Check review activity on the file.
- View Accepted vs. Rejected Findings

In the AI Review tab, you can find how many AI suggestions that are accepted, pending, and rejected.
- Accepted Findings: Shows all AI suggestions that were validated by a human reviewer.
- Rejected Findings: Displays all dismissed AI suggestions. These remain visible for historical context but do not clutter your active task manager.
- Pending Findings: Highlights any flagged items that still await human validation.
- Click a Historical Finding to View Location

Click on any past finding, whether accepted or rejected, kindly click the dropdowns in your AI review history. Every run is saved and visible to anyone with access to the PDF. Each run shows Run by, review type (Editorial Review / Custom Instructions), status (In progress / Completed / Failed) and counts (findings, accepted, accepted after edit, rejected, pending). Select a run to see its findings, each decision, and links to the tasks created. Starting a new review never hides earlier ones.
What These Status Tags Mean?
| Status Tag | What It Indicates | Task Association |
| Accepted | The finding was approved by a human reviewer. | zipBoard will create tasks under the user who accepted it and create comments with a spark icon. |
| Rejected | Finding was reviewed and explicitly declined. | No task created. Stored permanently in the history log for audit purposes. |
| Pending | Flagged during the latest Orbit Check review run. | Awaiting human validation in the AI Review tab. |
What does a Failed status mean, and what should I do if a review fails?
A Failed status means Orbit Check encountered an error while processing the document and could not complete the review. A review failure is typically caused by one of six reasons:
- Server Failure: High document complexity, such as long PDFs or temporary server overload.
- Network Error: A connection loss occurred between your browser and the review service while initiating or running the check.
- Credit Limit Reached: Your organization has exhausted its available AI review credit limit for the billing cycle.
- Rate Limits: An organization can run at most 10 AI reviews at the same time across different files and a single review at a time on the same file.
- Malicious or Invalid Prompt: Custom instructions contain restricted terms, policy-violating language, or instructions designed to bypass system safety controls.
- PDF not supported: This occurs when a file is too large or complex for Orbit Check to process. No credits are charged for failed attempts. Retrying the Run AI Review button do not resolve this issue, please split the PDF into smaller sections or contact support@zipboard.co.
What Happens When a Review Fails?
- If a review fails due to a server failure or network error, it will not count against your organization's AI credit limit. If a review fails due to invalid prompt, your credits are charged to read the document.
- All previous completed reviews, decisions, and existing tasks associated with the file remain completely safe and untouched in your AI Review history.
What You Should Do
Depending on the cause, follow these steps to resolve the failure:
- For Network Errors or Server Failures: Wait a moment and click Run AI Review to try again (or Re-run if you have already started a review round beforehand). For exceptionally long PDFs, try running focused custom instructions on specific sections. If you see the message "AI Review is temporarily disabled while we fix an issue. Please try again later," the feature has been temporarily paused for maintenance, and we'll have it back up shortly.
- For Credit Limits: Contact your admin to review your AI credit plan.
- For Rate Limits: Please wait for one of the runnings reviews to finish before starting a new one.
- For Malicious/Invalid Prompts: Reframe your Custom Instructions to focus on clear, standard document checks (e.g., proofreading, term consistency, or formatting rules) and remove any unusual system commands.
Still Stuck? If a review continues to fail on a standard file, reach out to zipBoard Support (support@zipboard.co) with the document details so our team can assist.
Frequently Asked Questions (FAQ)
Are rejected findings permanently saved?
Yes. Rejected findings are never deleted automatically. They remain logged in the AI Review tab so you can audit dismissed suggestions at any stage of the project lifecycle.
Can I see who accepted or rejected an AI finding?
Yes. Rejected findings show who rejected them and why, and each review shows who ran it.
Does running a new AI review overwrite past review history?
No. New Orbit Check review runs append new findings to the document's review log. You can review current findings alongside historical records for complete continuity.
Does Orbit Check store my document data?
Content is sent to AI services on Amazon Bedrock only when someone runs an AI feature. It isn't used to train models.