diff --git a/docs/content/federal_compliance/remediation_slas.md b/docs/content/federal_compliance/remediation_slas.md
index 691ee90013..7f67e6589f 100644
--- a/docs/content/federal_compliance/remediation_slas.md
+++ b/docs/content/federal_compliance/remediation_slas.md
@@ -36,10 +36,59 @@ threshold. **Internet-reachable** is signalled by a finding tag — `internet-re
 
 All the thresholds, tag names, and day counts are editable on the SLA configuration.
 
+Internet-reachability can also come from the computed asset exposure verdict rather than only a tag.
+Turn on **Use Asset Exposure for VDR Tiering** to include it. The two sources union, so enabling it
+can only tighten a deadline a tag already set, never loosen one.
+
 **FedRAMP VDR becomes mandatory on December 7, 2026.** FedRAMP's Vulnerability Detection and
 Response standard becomes mandatory for cloud service providers on that date. Adopting the VDR
 preset ahead of it is the recommended path.
 
+## Deadlines by Potential Agency Impact
+
+The three tiers above give every credibly exploitable, internet-reachable finding the same deadline,
+whatever the damage its exploitation would do. FedRAMP's published table does not: it crosses those
+same exploitability and reachability conditions with a **Potential Agency Impact N-rating** (PAIN),
+and the difference across ratings is large.
+
+Turn on **Use PAIN Ratings for VDR Deadlines** to switch from the three tiers to the full table:
+
+| PAIN rating | Exploitable **and** reachable | Exploitable only | Not credibly exploitable |
+| --- | --- | --- | --- |
+| N5 — debilitating effect on more than one agency | 2 days | 4 days | 16 days |
+| N4 — debilitating on one agency, or disruptive on several | 4 days | 8 days | 64 days |
+| N3 — disruptive effect on one agency | 16 days | 32 days | 128 days |
+| N2 — narrow customer effect | 48 days | 128 days | 192 days |
+
+Every cell is editable. The shipped numbers are FedRAMP's published Class C values; providers holding
+a Class B or Class D certification change the numbers, not the shape.
+
+### Rating your findings
+
+PAIN is set per finding, on the finding itself. It is deliberately a person's judgment rather than
+something computed from scanner output: FedRAMP asks the provider to estimate the effect exploitation
+would have on the agencies using the service, and explicitly declines to prescribe a method for
+arriving at that. A Triage Engine rule can propose a rating from evidence and route the finding for
+review — see the **Set Potential Agency Impact** node — but the confirmation is yours to make and
+defend.
+
+Two consequences worth knowing before you turn this on:
+
+* **A finding with no rating keeps its base deadline.** Every finding that exists before you start
+  rating is unrated, so nothing is re-dated the moment you enable this. Deadlines tighten as you rate.
+* **N1 has no row, on purpose.** FedRAMP's table starts at N2, so a finding rated N1 carries no VDR
+  deadline and keeps the FedRAMP Rev 5 window. DefectDojo does not invent a row FedRAMP has not
+  published.
+
+Enabling PAIN deadlines **replaces** the three tiers rather than combining with them. A finding rated
+N2 gets 48 days, not the 4-day tier it would have received without a rating — which is the point of
+rating it.
+
+The date each rating was set is recorded, because FedRAMP measures remediation timeframes from
+*evaluation* rather than from discovery, and asks reporting to show when each impact reduction
+happened. Re-running a rule that assigns the same rating a finding already has does not move that
+date.
+
 ## Relationship to the ledger
 
 SLA deadlines drive the scheduled completion dates on POA&M items, and determine which items count