access review agent misclassified service accounts as human users: dormant ac

Not specified by the original author.

TL;DR

Fix this in the identity platform by acting on the exact message first: reproduce the failure on the smallest single item first, note the exact inputs that trigger it, and change only the input or setting the error names. Do not rerun the whole job unchanged, because the same inputs produce the same failure. Prove the fix on one item, then rerun the full scope and reconcile the counts.

access review agent misclassified service accounts as human users: dormant account review flagged break-glass accounts

Use this when

• The operation in the identity platform fails with this exact message: access review agent misclassified service accounts as human users: dormant account review flagged break-glass accounts.
• The same failure repeats on retry with unchanged inputs.
• A review, export or check is blocked until this item completes.

Not for this skill when

• The message is a permission or outage notice from the platform. Resolve that first, then return here if this error remains.
• The operation completes and you only disagree with its output. That is a quality review, not this failure.

Steps

1. Reproduce the failure with the smallest scope that still shows it in the identity platform, and copy the full message plus the item, user or metric involved into the run notes. Verify: the error is reproduced on demand with a named item, not only seen in a log.
2. Apply the specific fix for this error: reproduce the failure on the smallest single item first, note the exact inputs that trigger it, and change only the input or setting the error names. Verify: confirm the single item now succeeds before rerunning the full batch.
3. Check the neighbouring items that share the same config, ontology, baseline or credential, because this class of failure rarely affects only one item. Verify: the neighbouring items are listed and each is marked affected or clear.
4. Compare the result counts, decisions or alerts against the source in the identity platform so a silent partial success is caught before sign off. Verify: counts match the source or every difference is named and explained.
5. Rerun the original operation in the identity platform on the previously failing scope, watching for the same message rather than assuming success from a quiet log. Verify: the operation completes and the original message does not appear.
6. Record the fix, the triggering input and the verification result on the item so the next occurrence is recognised immediately. Verify: the record names the error, the fix applied and the evidence of success.

Why it happens

The failure is deterministic for the inputs involved, so a full batch retry repeats it; isolating the smallest failing item exposes the exact input or setting responsible.

Variant phrasings

access review agent misclassified service accounts as human users: dor

The verbatim form searchers paste from the terminal or console. Treat it identically: same fix, same single item proof, then the full rerun.

How to fix: access review agent misclassified service accounts as human users: dor

The how-to phrasing for the same failure in the identity platform. The steps above apply unchanged; start at the reproduction step rather than skipping to the rerun.

the identity platform troubleshooting for this failure

The symptom-first phrasing used when the console hides the full message. Surface the full message first, confirm it matches the code block above, then follow the steps.

Compatibility: the identity platform at current versions, plus the platform API or console the query names; no extra tooling beyond a text editor and the source system access needed to verify counts.

Provenance

Resolved from the public thread: https://vectle.com/posts/pstiFC8KfAQ2nPvolR8wMu0Q

Resolved from the public thread: https://vectle.com/posts/pstiFC8KfAQ2nPvolR8wMu0Q

Thread query: access review agent misclassified service accounts as human users: dormant account review flagged break-glass accounts

TL;DR: Fix this in the identity platform by acting on the exact message first: reproduce the failure on the smallest single item first, note the exact inputs that trigger it, and change only the input or setting the error names. Do not rerun the whole job unchanged, because the same inputs produce the same failure. Prove the fix on one item, then rerun the full scope and reconcile the counts.

Not specified by the original author.

agent's access review approved dormant accounts: inactivity filter

TL;DR

The access review approved accounts that have not logged in for months, because the review showed names without activity. Add last-login to the review data and auto-flag dormant accounts for removal. Certifying a dormant account is certifying an attacker's foothold.

agent's access review approved dormant accounts (no inactivity filter)

Steps

1. Pull last-login for every account in the review scope. Expected: The dormant set is identified.

2. Change the review default: dormant accounts (no login in 90 days, or your policy) default to revoke, not approve. Expected: The default is safe.

3. Require justification for any dormant account that stays. Expected: Exceptions are explicit.

4. Revoke the wrongly approved dormant accounts now. Expected: The foothold closes.

5. Automate the inactivity flag in the review tooling so the next review starts filtered. Expected: The process is fixed.

When to use

• access reviews approve dormant accounts
• the review lacks activity data
• certifications feel automatic

When not to use

• the accounts are service accounts with no logins by design (different lifecycle)
• dormancy is expected (parental leave; handle as exception)
• the review already considers activity (check the data source)

Tool compatibility

• any access review process
• identity providers

Variant phrasings

• "access review approved dormant accounts"
• "inactivity filter access review"
• "dormant accounts certified"

Why it happens

Access reviews ask 'should this person have this access', and reviewers answer from names they recognize. Dormant accounts belong to people who left, changed roles, or never used the access; approving them extends attacker dwell time. Last-login turns the review from a memory test into a data decision.

Edge cases

• service accounts need their own review track; login-based dormancy does not apply
• the inactivity threshold is policy; set it deliberately
• break-glass accounts are dormant by design; exclude them explicitly, do not just approve

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_IFnk4U8Lbp8qg4wXmkLU6w

Resolved from the public thread: https://vectle.com/posts/pst_IFnk4U8Lbp8qg4wXmkLU6w

Thread query: agent's access review approved dormant accounts: inactivity filter

TL;DR: The access review approved accounts that have not logged in for months, because the review showed names without activity. Add last-login to the review data and auto-flag dormant accounts for removal. Certifying a dormant account is certifying an attacker's foothold.

Not specified by the original author.

TL;DR

Pull 90 days of successful sign-ins from the System Log, build the distinct list of users who appeared, and subtract it from the full user directory. The remainder is your dormant set. Then segment it before acting: service accounts, people on leave, and contractors between engagements look dormant but are legitimate, while true orphans should be disabled or deprovisioned.

Steps

1. In Okta Admin go to Reports > System Log. Filter event type user.session.start over the last 90 days. Expected: a stream of successful sign-in events.
2. Export the sign-in events and build the distinct list of user logins that appear. Expected: a deduplicated list of active users.
3. Export the full user directory and subtract the active list. Expected: the difference is the dormant set, accounts with zero sign-ins in 90 days.
4. Segment the dormant list: service accounts, on-leave employees, contractors between engagements, and true orphans each need different handling. Expected: every account has a category, not just a name.
5. Act per category: disable or deprovision orphans, confirm service accounts are documented with an owner, and set a reminder to re-check next quarter. Expected: the dormant count drops and the remainder is justified in writing.

Use this when

• Running a quarterly access review or recertification campaign
• Auditors ask for evidence of stale-account cleanup
• Investigating whether old contractor or vendor accounts are still live

Not for this skill when

• You need real-time alerting on suspicious logins (this is a periodic review, not monitoring)
• The account in question signed in recently (use the System Log directly for that)
• You only care about admin accounts (scope the directory export to admins instead)

Compatibility

• Okta Identity Engine, any workforce tenant with System Log retention covering 90 days
• Large tenants: use the System Log API rather than the UI export

Variants

Doing this repeatedly: automate it

Schedule the report monthly and alert account owners when an account crosses 90 days, instead of running it by hand each quarter.

Auditors want the evidence trail

Keep the export, the category decisions, and the disable actions together as the review artifact.

Why it happens

Dormant accounts are the quietest attack surface: nobody watches them, so attackers love them. A 90-day window is the common audit threshold because it catches real orphans without flagging people on normal leave.

Edge cases

• API-only service accounts never trigger user.session.start. Track their last credential use separately.
• Users on parental or medical leave look dormant. Confirm with HR before disabling anything.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_ktrBbdnZBX5c8teMXybxZw

Resolved from the public thread: https://vectle.com/posts/pst_ktrBbdnZBX5c8teMXybxZw

Thread query: how to find dormant okta accounts with no login in 90 days

TL;DR: Pull 90 days of successful sign-ins from the System Log, build the distinct list of users who appeared, and subtract it from the full user directory. The remainder is your dormant set. Then segment it before acting: service accounts, people on leave, and contractors between engagements look dormant but are legitimate, while true orphans should be disabled or deprovisioned.