AI-Drafted Reviews
Prepare AI review drafts from available goals, prior reviews, and work signals, then check and authorize the content before submission.
Why It Matters
The blank page is why reviews are late, and recall bias is why they’re unfair. Kit’s AI drafter attacks both: it gathers real evidence (the reviewee’s goals and check-ins, their previous evaluation, their role expectations, their own notes, and (opt-in) their shipped work) and produces a structured draft answering the cycle’s questions. Generating a draft does not submit it.
Before submission, the reviewer must check the draft, correct errors, and justify the assessment. Kit records AI assistance and either a content change or explicit ownership. These checks do not verify the quality of human review or establish legal compliance.
Important
The human reviewer is accountable for every submitted review. An AI-assisted review cannot be submitted until the reviewer has changed the draft content or explicitly checked “I have reviewed this AI draft and take ownership of it as my own assessment.” Both paths are timestamped into the review’s metadata. An authorized MCP assistant can submit your own review on your instruction, subject to submission checks. You remain responsible for reviewing and approving its content.
Generating a Draft
Open any review in your queue and press Draft with AI. Generation is synchronous; a few seconds later the form is filled in and ready to edit. The available evidence depends on the reviewer role. Managers and self-reviewers can use all sources below. Peer reviewers receive only role expectations and optional work signals.
| Evidence | Source |
|---|---|
| Goals + check-ins | The reviewee’s goals: title, status, progress, latest note |
| Previous evaluation | The manager summary from the reviewee’s last finalized evidence record |
| Role expectations | The participant’s role summary for this cycle |
| The reviewee’s own notes | The Your own notes field from their self-review: tool-invisible work like mentoring and incidents |
| Work signals (opt-in) | Merged MRs and review activity fetched live from connected sources, only if you tick the per-draft checkbox |
The output fills the cycle’s question answers, a proposed overall rating (only if it fits the template’s scale), and a summary written as themes plus representative contributions. The review is flagged AI-assisted from that moment, and the flag follows it into the evidence record.
The prompt is narrative-only by construction: the drafter cites representative work, credits AI-direction and AI-enablement contributions, and is forbidden from producing volume metrics, rankings, or cross-person comparisons.
The Accountability Trail
Draft metadata is encrypted and stored on the review itself:
- Generation: when it was drafted, who requested it, and a verbatim copy of what the AI produced.
- Content change: saves compare answers, summary, and overall rating against the stored draft. The first difference records an edit timestamp; Kit does not assess how substantial that change is.
- Explicit ownership: submitting an unedited draft requires the ownership checkbox, which stamps an ownership timestamp and your name.
- Signal disclosure: if work signals were used, the used items are retained on the review and shown to its reviewer. The reviewee does not automatically have access to this panel in someone else’s review; discuss sources with them before finalization.
The finalized evaluation record copies review content and the AI-assisted flag, not the full draft metadata or signal list. It flags each review as AI-assisted or not, and the evidence PDF prints it as AI-assisted draft: Yes/No for the reviewee and auditor.
| Attempted action | What Kit does |
|---|---|
| Submit an untouched AI draft, checkbox unticked | Refused: “must be edited or explicitly accepted as your own assessment before submitting” |
| Submit after content changes | Allowed; edit timestamp in the trail |
| Submit unedited with the ownership box ticked | Allowed; ownership timestamp + name in the trail |
| Generating a draft | Saves without submission; a separate MCP tool can submit the acting user’s own review when submission conditions are met |
Credits and Failure Modes
Drafting runs on your account’s AI credits, metered like every other AI feature (see Integrations > AI Settings), one LLM completion per draft. If credits are exhausted you get a dedicated message (“This account has no AI credits left. Write the review yourself or add credits.”) rather than a silent failure. If the model or a signal source is unreachable, the draft fails cleanly or degrades (a dead GitLab becomes a note in the draft context, never a broken form) and you can always write by hand.
Quick Checklist
- Ask reviewees to tend goals and Your own notes before you draft
- Press Draft with AI; tick the work-signals box only with the reviewee’s awareness
- Edit the draft; verify every claim against your own judgment
- If you change nothing, tick the ownership checkbox
- Check the signal disclosure panel shows what you expect before submitting
Next Steps
- Work Signal Sources: connect GitLab so drafts cite real shipped work
- SOC 2 Evidence and Exports: how the AI-assisted flag appears to an auditor