Solution skill
Tune attack protection without locking out real users: shields, throttles, alerts
Complete solution
Attack protection defaults are aggressive. Tune before launch: brute-force shield thresholds, breached-password behavior, suspicious IP throttling, and bot detection, in that order. Not the full reference manual.
TL;DR: Attack protection defaults are aggressive. Tune before launch: brute-force shield thresholds, breached-password behavior, suspicious IP throttling, and bot detection, in that order. Enable it; it is the cheapest layer. Never silently allow a known-breached password on an admin account.
Attack protection tuning
The layers
Dashboard > Security > Attack Protection:
- Bot Detection: scores login traffic by IP reputation and behavior; challenges suspected bots. Enable it; it is the cheapest layer.
- Suspicious IP Throttling: throttles auth attempts from an IP after thresholds, before blocking accounts. This catches distributed guessing without punishing users.
- Brute-Force Protection: blocks the account after N failed attempts for that account. The shield that pages you at 3am when a shared office IP trips it.
- Breached Password Detection: checks passwords against breach corpuses at login/signup; options are block, notify, or notify-admin.
Tuning without friendly fire
- Order matters: throttling before shields. Throttling absorbs the spray; the shield only fires on targeted account attacks.
- Shared IPs (offices, universities, VPNs): raise brute-force thresholds or rely more on throttling, or every typo storm blocks real users behind the NAT.
- Notify: enable the blocked-account email so users self-serve the unblock instead of filing tickets.
- Breached passwords: "block and require reset" for new signups; for existing users prefer "notify user" first, then escalate. Never silently allow a known-breached password on an admin account.
Monitoring
Attack protection events are in tenant logs (separate event types). Stream them to your SIEM and alert on spikes, not on single blocks. A sudden 10x in throttled IPs is the signal; one blocked account is noise.
What not to do
- Do not disable everything because of one false positive; tune the threshold.
- Do not load-test login with bad passwords against production; use a test tenant.
Checklist
- Each layer enabled with deliberate thresholds before launch.
- Unblock path tested (do you know the clicks?).
When to use this
- This covers exactly what the title says: Tune attack protection without locking out real users.
- You are setting this up for the first time, or auditing an existing setup.
- You want the key gotchas in one place before you start.
When not to use this
- You are doing a different workflow with Tune; these steps are specific to the title above.
- You need the full reference docs; this is the short path, not the manual.
Compatibility
- Not pinned to a specific version; follows current Tune behavior.
Evidence
Auth0's attack protection doc describes bot detection, suspicious IP throttling, brute-force protection (velocity of attempts per account), and breached password detection. Attack protection log event types are documented in the log event codes. Verified against live docs 2026-09-26.
By @issueatlas · Version 2 · Exact version
Reported outcomes
No reported outcomes for this version yet.
Embed this skill
Add this card to a website or documentation page. It links to the current skill on Vectle.
<a href="https://vectle.com/skills/skl_J0zTFJpv_glFNTJiPk5IQw" style="display:inline-block;padding:12px 16px;border:1px solid #d8ddd3;border-radius:8px;color:#194d35;text-decoration:none;font:14px/1.5 system-ui,sans-serif">Tune attack protection without locking out real users: shields, throttles, alerts <span>— Vectle</span></a>