> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mergeoa.com/llms.txt
> Use this file to discover all available pages before exploring further.
# The Candidate Experience During a Merge Assessment
> See exactly what candidates encounter in a Merge assessment, from the invite email to the codebase browser and PR review interface.
This page walks through what candidates see and do during a Merge assessment. Understanding the candidate experience helps you set accurate expectations in your outreach, prepare candidates to perform their best, and interpret results in the context of how the session actually unfolds.
## Receiving the invite
Candidates receive a unique email link from your hiring organization. The link is single-use and tied to their email address, so it cannot be forwarded or shared.
When they click the link, candidates enter a pre-assessment setup flow before the timer starts. The setup flow covers:
* **Format overview** — a plain-language explanation of the review loop, what candidates are expected to do, and what the AI revision means
* **Time expectations** — the session length and what happens when time expires
* **Proctoring requirements** — if you have enabled proctoring, candidates are shown the specific rules before they agree to proceed
* **Identity verification** — candidates must confirm their identity before the assessment begins
Nothing in the setup flow is timed. Candidates can read at their own pace and ask clarifying questions before committing to start.
## The assessment interface
Once candidates start the session, they see a split-pane interface. The codebase browser sits on the left; the PR diff sits on the right.
Candidates can navigate files freely, read the commit history, and build context for the broader codebase — not just the changed lines. This mirrors how a real reviewer approaches an unfamiliar PR, and it's intentional: surface-level diff reading is not the same as understanding the change.
Inline commenting works like a standard code review tool. Candidates click a line, write a comment, and optionally tag it:
* **Bug** — a correctness issue that should block merge
* **Refactor** — a structural improvement that doesn't change behavior
* **Security** — a vulnerability or unsafe pattern
* **Style** — a formatting or readability concern
Tags are optional but help candidates organize their thinking and make the review easier to interpret.
## Submitting comments and iterating
Once candidates are satisfied with their initial comments, they submit the review. Merge's AI processes the comments and opens a revision within seconds — no waiting, no manual steps.
The updated diff appears in the same interface. Candidates continue reviewing the new revision exactly as they reviewed the first: reading the changes, leaving inline comments, and deciding what still needs to be addressed.
The loop ends when the candidate clicks **Approve** or the session timer expires. There is no fixed number of revision rounds — the loop runs as many times as the session window allows.
## Completing the assessment
When the session ends, candidates see a completion screen that confirms their submission was received. The screen thanks them for their time and explains what happens next in your hiring process.
Candidates do not see their scorecard. Results — including comment quality, revision response, and overall assessment signals — are sent to your hiring team only. This keeps the evaluation objective and consistent across candidates.
Candidates should use their own computer and browser for the assessment. Switching tabs or windows during the session may be flagged if proctoring is enabled. Share this with candidates in advance so they can close distractions before they start.
Encourage candidates to treat this like a real PR review — focus on the most impactful comments, not exhaustive line-by-line notes. The best reviewers are selective: they surface the issues that matter most and explain why, rather than cataloging every possible observation.