> ## Documentation Index
> Fetch the complete documentation index at: https://help.liftai.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Score a contract with ContractIQ

> Run an AI review that rates each contract clause as favorable, neutral, or unfavorable from the owner's side, then publish a client-facing report.

**ContractIQ** is an AI-assisted review that reads an elevator service contract from the property owner's perspective. It locates each clause in the attached PDF, rates it against Lift AI's consultant rubric, and drafts a report you can edit and share with the client. The ContractIQ tab sits alongside Overview, Coverage, Rates, KPIs, and Attachments in the contract detail panel.

<Note>
  **ContractIQ is in beta.** Scores and findings are still being refined and may change. We'd love your feedback — send it to us through the **Support** button in the sidebar. In the app, the ContractIQ tab is marked **Beta** and the same note appears above each report.
</Note>

<Note>
  **Consultants** run, edit, and publish ContractIQ reports. **Client** users see a report only after it's published.
</Note>

## What ContractIQ scores

A run scores **18 core provisions** — term, termination, price escalation, scope, liability, and so on — and flags additional noteworthy clauses it discovers in the document. Each finding gets:

* A **rating**: **Favorable**, **Neutral**, **Unfavorable**, or **Not applicable**.
* A short **summary** written for the report.
* A **recommended action** for clauses that are below market or unfavorable.
* The **contract text it relied on**, so you can verify every finding against the source.

A **Findings summary** card tallies the findings by rating, and an **Overall assessment** gives the owner a plain-language verdict on the agreement.

<h2 id="run-a-report">
  Run a report
</h2>

<Steps>
  <Step title="Attach the contract PDF">
    ContractIQ reads the contract document, so the contract needs a PDF on its **Attachments** tab. Until one is attached, the run button is disabled with the note **Attach the contract PDF first**.
  </Step>

  <Step title="Open the ContractIQ tab and run">
    On the contract, open the **ContractIQ** tab and select **Run ContractIQ**. A run usually finishes in a few minutes — you can leave the tab open or come back later, since the report is saved when it finishes.
  </Step>

  <Step title="Wait for scoring">
    While the run is in progress, the tab shows its status and keeps refreshing on its own. If a run fails, an error explains why and a **Retry** button reruns it.
  </Step>
</Steps>

When scoring finishes, the report opens for your review.

## Review and edit the findings

A finished report opens in the editor, where you can adjust anything before clients see it.

* **Change a rating** — each finding has a rating you can override. The summary counts update as you go.
* **Edit the summary and recommended action** — both are prepopulated by the agent; edit them to override. The summary and the recommended action are what appear in the client-facing report.
* **Check the evidence** — expand **Show AI original & evidence** on a finding to see the agent's original rating, full rationale, recommended action, and the quoted contract language it cited, along with a confidence percentage.
* **Edit the Overall assessment** — write the owner-facing verdict at the top of the report. Clearing the text reverts to the AI-generated version.
* **Drop a finding** — remove a finding from the report with the trash icon (**Drop from report**); use **Undo** to bring it back. Dropped findings don't appear in the summary or the published report.
* **Add a custom finding** — select **Add custom finding** to add your own, with a title, a rating, and a rationale. Custom findings are attributed to you.

Each finding is badged by source — **Core provision**, **AI-discovered**, or **Custom** — so you can tell what the agent scored from what you added. When the agent's rating conflicts with prior consultant precedent, a **Rubric / precedent conflict** note flags the finding for closer review.

## Preview and publish

A footer at the bottom of the editor shows the running rating counts and two actions:

* **Preview report** opens the current draft as a report in a new browser tab, so you can read exactly what the client will see before committing. Save it as a PDF from your browser's print dialog if you want a copy.
* **Publish report** finalizes the report and makes it visible to the client. There's no separate approve step — publishing is the single finalize action.

After publishing, the tab shows when it was published and a **View report** button. Only published reports are used to inform future runs on other contracts.

## Reopen a published report

To change a published report, select **Reopen for edits**. The report returns to the editing state so you can make changes and publish again; the republished report notes that it was re-published after edits.

## What clients see

Client users never see runs in progress or work-in-progress drafts. On the ContractIQ tab they see a **View report** button once a report is published, or a note that no report has been published yet. Opening the report shows the same findings, summary, and overall assessment, read-only.

## Related

* [View contract details](/contracts/view-a-contract)
* [How contracts power invoice AI review](/contracts/contracts-and-ai-review)
* [Edit a contract](/contracts/edit-a-contract)