Review Glean Assistant security posture
Use Assistant in Protect to see how much risk Glean Assistant carries and which runtime issues drive it. The page combines the Glean Assistant risk score and summary with its recent issues and the tools enabled for it.
Prerequisites
Complete the following prerequisites:
- Confirm that agent posture is enabled for your environment. If Security posture doesn't appear in Protect, contact your Glean account team.
- Have the Super Admin or Sensitive Content Moderator role. See Administrator Roles.
Open the Assistant posture page
To review Glean Assistant:
- Open Protect from the Glean navigation.
- Under Security posture, click Assistant.
Interpret the score
The Risk summary shows the Glean Assistant score out of 100, along with a written summary of its risks. The score uses the same risk levels as agents:
| Risk level | Score |
|---|---|
| Low | 0–24 |
| Medium | 25–49 |
| High | 50–74 |
| Critical | 75–100 |
Glean Assistant doesn't have a configuration scan of its own, so its score comes from two inputs:
- Open runtime issues from the past 10 days make up 90% of the score. More severe issues count for more, so one critical issue outweighs two high issues. Closing an issue lowers the score.
- The scan risk of the enabled tools makes up the other 10%. Glean averages the scan scores of up to five of the riskiest scanned tools. Tools without a current scan don't contribute, and this part counts as 0 when no enabled tool has been scanned.
Glean recalculates the score hourly. It rewrites the summary only when the underlying data changes, and at most every six hours, so the summary can trail the score. A summary will be available after the next scan means that Glean hasn't written a summary yet.
Unassessed means that Glean hasn't calculated a current score for Glean Assistant yet.
The Glean Assistant score also counts for 30% of the overall risk score on Overview. See Review overall risk posture.
Review issues
The Issues trend chart shows how many issues Glean detected for Glean Assistant each day over the past 10 days, by severity. It counts every detected issue, including issues that were closed later. Hover over a day to see its counts.
Open issues lists the latest open issues, newest first, with each issue's title, enforcement action when present, and detection time. The list shows up to 10 issues. When it reaches that limit, Showing the latest 10 open issues appears and older open issues aren't listed.
Click an issue to open its details pane, the same pane that AI security issues uses. Triage the issue there, or move to the next or previous issue in the list. When you close the pane, the list refreshes. See Findings Dashboard.
Review enabled tools
Other details shows the Last scanned time, which is when Glean last wrote the risk summary, and the tools enabled for Glean Assistant. The tools are counted as write tools and read tools, and the page notes how many write tools can run automatically. No tools enabled means that Glean Assistant has no enabled tools.
The tool counts give context for the score but don't change it. Only the scan risk of scanned tools affects the score.
See also
- Review agent security posture — investigate the risk scores and findings of individual agents
- Configuring Policies — manage the guardrails that apply to Glean Assistant and agents