diff --git a/jmp.sh b/jmp.sh
deleted file mode 100755
index 7056955..0000000
--- a/jmp.sh
+++ /dev/null
@@ -1,7 +0,0 @@
-#!/bin/bash
-
-# get some jmp.chat numbers to feed phone-mnemonic one of these two ways:
-
-curl 'https://jmp.chat/sp1a/register1/rc.php?state=ky&city=louisville' | grep -Eo '\([0-9]{3}\) [0-9]{3}-[0-9]{4}' | tr -dc '[0-9\n]' >> nums.txt
-
-curl 'https://jmp.chat/sp1a/register1/?areacode=289' | grep -Eo '[0-9]{11}' >> nums.txt
diff --git a/jmp/jmp.sh b/jmp/jmp.sh
new file mode 100755
index 0000000..0760f92
--- /dev/null
+++ b/jmp/jmp.sh
@@ -0,0 +1,49 @@
+#!/bin/bash
+set -uo pipefail
+
+# run this script to download all available jmp.chat numbers
+
+# get some jmp.chat numbers to feed phone-mnemonic one of these two ways manually:
+
+# curl 'https://jmp.chat/sp1a/register1/rc.php?state=ky&city=louisville' | grep -Eo '\([0-9]{3}\) [0-9]{3}-[0-9]{4}' | tr -dc '[0-9\n]' >> nums.txt
+
+# curl 'https://jmp.chat/sp1a/register1/?areacode=289' | grep -Eo '[0-9]{11}' | sed 's/^1//' >> nums.txt
+
+# rm -rf us rate-centers
+mkdir -p us rate-centers
+
+# get US area codes once:
+[ ! -f areacodes-us.txt ] && curl -s https://www.worldatlas.com/na/us/area-codes.html | grep -Eo 'area-code-[0-9]{3}\.html' | grep -Eo '[0-9]{3}' | sort -un > areacodes-us.txt
+
+# first get all by area code, but this API is limited to 5000 in a single response
+while IFS="" read -r area_code || [ -n "$area_code" ]
+do
+ # this API has the leading 1, strip it off
+ echo $area_code
+ curl -s "https://jmp.chat/sp1a/register1/?areacode=$area_code" | grep -Eo '[0-9]{11}' | sed 's/^1//' | sort -un > "us/$area_code.txt"
+ sleep 1
+done < areacodes-us.txt
+
+# get US states once:
+[ ! -f states.txt ] && curl -s 'https://worldpopulationreview.com/static/states/abbr-list.csv' | tail -n+2 | tr -d '"' > states.txt
+
+# now get all by rate center, grouped by state
+while IFS="" read -r state || [ -n "$state" ]
+do
+ # get rate centers, but only once
+ [ ! -f "rate-centers/$state.txt" ] && curl -s "https://jmp.chat/sp1a/register1/all_rc.php?state=$state" | grep -Eo '[^<]+' | sed -e 's@^@@' -e 's@$@@' | sort -u > "rate-centers/$state.txt" && sleep 1
+
+ rm -f "us/$state.txt"
+
+ while IFS="" read -r city || [ -n "$city" ]
+ do
+ echo "$state / $city"
+ # this API does not have the trailing 1, but uses a different format, only keep numbers
+ curl -s "https://jmp.chat/sp1a/register1/rc.php?state=$state&city=$city" | grep -Eo '\([0-9]{3}\) [0-9]{3}-[0-9]{4}' | tr -dc '[0-9\n]' >> "us/$state.txt"
+ sleep 1
+ done < "rate-centers/$state.txt"
+
+ # de-duplicate
+ [ -f "us/$state.txt" ] && sort -un "us/$state.txt" | sponge "us/$state.txt"
+
+done < states.txt
\ No newline at end of file