diff --git a/.github/PULL_REQUEST_TEMPLATE.md b/.github/PULL_REQUEST_TEMPLATE.md
new file mode 100644
index 00000000000..36292805d5b
--- /dev/null
+++ b/.github/PULL_REQUEST_TEMPLATE.md
@@ -0,0 +1,5 @@
+<!-- Write **BELOW** The Headers and **ABOVE** The comments else it may not be viewable. -->
+
+## About The Pull Request
+
+<!-- Describe The Pull Request. -->
diff --git a/tools/GenerateChangelog/How to update the changelogs.txt b/tools/GenerateChangelog/How to update the changelogs.txt
deleted file mode 100644
index 51cd9919c45..00000000000
--- a/tools/GenerateChangelog/How to update the changelogs.txt	
+++ /dev/null
@@ -1,15 +0,0 @@
-Taken mostly from https://tgstation13.org/wiki/Guide_to_Changelogs#Updating_the_Changelog with a few changes.
-
-First, install Python 2.7. Yes, it's outdated, but that's what the script is made to work with.
-Make sure that Python 2.7 and its Scripts/ folder is in PATH
-Upgrade pip, as outlined here https://pip.pypa.io/en/stable/installing/
-Once pip is installed, you will need to run two commands from your terminal or command prompt to install packages
-1. 'pip install PyYaml'
-2. 'pip install beautifulsoup4'
-
-Once all this is done, you should just be able to run ss13_genchangelog.py and sit back.
-
-If you get errors running it, two things are possible.
-1. You installed something wrong.
-2. Someone formatted their changelog wrong.
-Check where the error was, do a bit of troubleshooting, see if it wasn't #2. Fixing the syntax in the yaml files is pretty simple.
\ No newline at end of file
diff --git a/tools/GenerateChangelog/makeChangelog.bat b/tools/GenerateChangelog/makeChangelog.bat
deleted file mode 100644
index 7ee7bb625da..00000000000
--- a/tools/GenerateChangelog/makeChangelog.bat
+++ /dev/null
@@ -1,4 +0,0 @@
-@echo off
-rem Cheridan asked for this. - N3X
-call python ss13_genchangelog.py ../../html/changelog.html ../../html/changelogs
-pause
diff --git a/tools/GenerateChangelog/ss13_autochangelog.py b/tools/GenerateChangelog/ss13_autochangelog.py
deleted file mode 100644
index 844f7050dd7..00000000000
--- a/tools/GenerateChangelog/ss13_autochangelog.py
+++ /dev/null
@@ -1,121 +0,0 @@
-'''
-Usage:
-    $ python ss13_autochangelog.py [--dry-run] html/changelogs [PR_number] [PR_author] [PR_Body]
-
-ss13_autochangelog.py - Generate changelog YAML files from pull request.
-'''
-
-from __future__ import print_function
-import os, sys, re, time, argparse
-from datetime import datetime, date
-from time import time
-
-today = date.today()
-
-dateformat = "%d %B %Y"
-
-opt = argparse.ArgumentParser()
-opt.add_argument('target_dir', help='The location to write changelog files to.')
-opt.add_argument('pr_numb', help='The number of the pull request.')
-opt.add_argument('pr_author', help='The author of the pull request. If specific authors aren\'t specified, the PR author will be used')
-opt.add_argument('pr_body', help='The body of the pull request to parse for changelogs.')
-
-args = opt.parse_args()
-
-all_changelog_entries = {}
-
-validPrefixes = {
-    "fix": 'bugfix',
-	"fixes": 'bugfix',
-	"bugfix": 'bugfix',
-    "wip": 'wip',
-    "tweak": 'tweak',
-	"tweaks": 'tweak',
-	"rsctweak": 'tweak',
-    "soundadd": 'soundadd',
-    "sounddel": 'sounddel',
-    "add": 'rscadd',
-    "adds": 'rscadd',
-	"rscadd": 'rscadd',
-    "del": 'rscdel',
-	"dels": 'rscdel',
-	"delete": 'rscdel',
-	"deletes": 'rscdel',
-	"rscdeldel": 'rscdel',
-	"imageadd": 'imageadd',
-    "imagedel": 'imagedel',
-    "maptweak": 'maptweak',
-	"remap": 'maptweak',
-	"remaps": 'maptweak',
-    "typo": 'spellcheck',
-	"spellcheck": 'spellcheck',
-    "experimental": 'experiment',
-	"experiments": 'experiment',
-	"experiment": 'experiment'
-}
-
-incltag = False
-new_logs = {}
-author = args.pr_author
-new = 0
-
-# Parse PR body for changelog entries
-print('Reading changelogs...')
-for line in args.pr_body.splitlines():
-	print(f"Checking line '{line}'")
-	if line[:1] == "🆑": # Find the start of the changelog
-		print("Found opening :cl: tag")
-		if incltag == True: # If we're already reading logs, skip
-			continue
-		incltag = True
-
-		# Fetch the author name
-		author = line[1:]
-		author.strip()
-
-		if not len(author):
-			author = args.pr_author
-
-		if author not in new_logs:
-			new_logs[author] = [] # Make array entry for the author
-		continue
-
-	# If we hit a /cl, we're no longer reading logs
-	elif line == "/🆑":
-		print("Found closing /:cl: tag")
-		incltag = False
-
-	# If we aren't reading logs, we don't care about any other line contents
-	if not incltag:
-		continue
-
-	# Split line into tag (icon) and body (comment)
-	body = re.split("[ ,:-]", line, 1)
-	if len(body) != 2:
-		continue # If there's just one word, then it can't really be a changelog, now can it
-
-	if body[0] in validPrefixes:
-		tag = validPrefixes[body[0]]
-		body = body[1].strip(" ,-:\t\n")
-	else: # If the tag is invalid, just default to rscadd
-		tag = "rscadd"
-		body = line.strip(" ,-:\t\n")
-
-	new_logs[author].append(f"  - {tag}: \"{body}\"")
-	new += 1
-
-print(f"Writing {new} new changelog entries.") # f supposedly formats new into the var
-
-for auth in new_logs:
-	# Sanitize authors without changes
-	if not len(new_logs[auth]):
-		continue
-
-	f = open(os.path.join(args.target_dir, f"{auth}{args.pr_numb}.yml"), 'w')
-	print(f"Writing changes to {f}")
-	f.write(f'author: {auth}\n')
-	f.write('delete-after: True\n')
-	f.write('changes:\n')
-	for log in new_logs[auth]:
-		f.write(f'{log}\n')
-	f.close()
diff --git a/tools/GenerateChangelog/ss13_genchangelog.py b/tools/GenerateChangelog/ss13_genchangelog.py
deleted file mode 100644
index e597ac7a6c5..00000000000
--- a/tools/GenerateChangelog/ss13_genchangelog.py
+++ /dev/null
@@ -1,215 +0,0 @@
-'''
-Usage:
-    $ python ss13_genchangelog.py [--dry-run] html/changelog.html html/changelogs/
-
-ss13_genchangelog.py - Generate changelog from YAML.
-
-Copyright 2013 Rob "N3X15" Nelson <nexis@7chan.org>
-
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in
-all copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
-THE SOFTWARE.
-'''
-
-from __future__ import print_function
-import yaml, os, glob, sys, re, time, argparse
-from datetime import datetime, date
-from time import time
-
-today = date.today()
-
-dateformat = "%d %B %Y"
-
-opt = argparse.ArgumentParser()
-opt.add_argument('-d', '--dry-run', dest='dryRun', default=False, action='store_true', help='Only parse changelogs and, if needed, the targetFile. (A .dry_changelog.yml will be output for debugging purposes.)')
-opt.add_argument('targetFile', help='The HTML changelog we wish to update.')
-opt.add_argument('ymlDir', help='The directory of YAML changelogs we will use.')
-
-args = opt.parse_args()
-
-all_changelog_entries = {}
-
-validPrefixes = [
-    'bugfix',
-    'wip',
-    'tweak',
-    'soundadd',
-    'sounddel',
-    'rscdel',
-    'rscadd',
-    'imageadd',
-    'imagedel',
-    'maptweak',
-    'spellcheck',
-    'experiment'
-]
-
-def dictToTuples(inp):
-    return [(k, v) for k, v in inp.items()]
-
-changelog_cache = os.path.join(args.ymlDir, '.all_changelog.yml')
-
-failed_cache_read = True
-if os.path.isfile(changelog_cache):
-    try:
-        with open(changelog_cache) as f:
-            (_, all_changelog_entries) = yaml.load_all(f)
-            failed_cache_read = False
-
-            # Convert old timestamps to newer format.
-            new_entries = {}
-            for _date in all_changelog_entries.keys():
-                ty = type(_date).__name__
-                # print(ty)
-                if ty in ['str', 'unicode']:
-                    temp_data = all_changelog_entries[_date]
-                    _date = datetime.strptime(_date, dateformat).date()
-                    new_entries[_date] = temp_data
-                else:
-                    new_entries[_date] = all_changelog_entries[_date]
-            all_changelog_entries = new_entries
-    except Exception as e:
-        print("Failed to read cache:")
-        print(e, file=sys.stderr)
-
-if args.dryRun:
-    changelog_cache = os.path.join(args.ymlDir, '.dry_changelog.yml')
-
-if failed_cache_read and os.path.isfile(args.targetFile):
-    from bs4 import BeautifulSoup
-    from bs4.element import NavigableString
-    print(' Generating cache...')
-    with open(args.targetFile, 'r') as f:
-        soup = BeautifulSoup(f)
-        for e in soup.find_all('div', {'class':'commit'}):
-            entry = {}
-            date = datetime.strptime(e.h2.string.strip(), dateformat).date()  # key
-            for authorT in e.find_all('h3', {'class':'author'}):
-                author = authorT.string
-                # Strip suffix
-                if author.endswith('updated:'):
-                    author = author[:-8]
-                author = author.strip()
-
-                # Find <ul>
-                ulT = authorT.next_sibling
-                while(ulT.name != 'ul'):
-                    ulT = ulT.next_sibling
-                changes = []
-
-                for changeT in ulT.children:
-                    if changeT.name != 'li': continue
-                    val = changeT.decode_contents(formatter="html")
-                    newdat = {changeT['class'][0] + '': val + ''}
-                    if newdat not in changes:
-                        changes += [newdat]
-
-                if len(changes) > 0:
-                    entry[author] = changes
-            if date in all_changelog_entries:
-                all_changelog_entries[date].update(entry)
-            else:
-                all_changelog_entries[date] = entry
-
-del_after = []
-errors = False
-print('Reading changelogs...')
-for fileName in glob.glob(os.path.join(args.ymlDir, "*.yml")):
-    name, ext = os.path.splitext(os.path.basename(fileName))
-    if name.startswith('.'): continue
-    if name == 'example': continue
-    fileName = os.path.abspath(fileName)
-    print(' Reading {}...'.format(fileName))
-    cl = {}
-    with open(fileName, 'r') as f:
-        cl = yaml.load(f)
-        f.close()
-    if today not in all_changelog_entries:
-        all_changelog_entries[today] = {}
-    author_entries = all_changelog_entries[today].get(cl['author'], [])
-    if len(cl['changes']):
-        new = 0
-        for change in cl['changes']:
-            if change not in author_entries:
-                (change_type, _) = dictToTuples(change)[0]
-                if change_type not in validPrefixes:
-                    errors = True
-                    print('  {0}: Invalid prefix {1}'.format(fileName, change_type), file=sys.stderr)
-                author_entries += [change]
-                new += 1
-        all_changelog_entries[today][cl['author']] = author_entries
-        if new > 0:
-            print('  Added {0} new changelog entries.'.format(new))
-
-    if cl.get('delete-after', False):
-        if os.path.isfile(fileName):
-            if args.dryRun:
-                print('  Would delete {0} (delete-after set)...'.format(fileName))
-            else:
-                del_after += [fileName]
-
-    if args.dryRun: continue
-
-    cl['changes'] = []
-    with open(fileName, 'w') as f:
-        yaml.dump(cl, f, default_flow_style=False)
-
-targetDir = os.path.dirname(args.targetFile)
-
-with open(args.targetFile.replace('.htm', '.dry.htm') if args.dryRun else args.targetFile, 'w') as changelog:
-    with open(os.path.join(targetDir, 'templates', 'header.html'), 'r') as h:
-        for line in h:
-            changelog.write(line)
-
-    for _date in reversed(sorted(all_changelog_entries.keys())):
-        entry_htm = '\n'
-        entry_htm += '\t\t\t<h2 class="date">{date}</h2>\n'.format(date=_date.strftime(dateformat))
-        write_entry = False
-        for author in sorted(all_changelog_entries[_date].keys()):
-            if len(all_changelog_entries[_date]) == 0: continue
-            author_htm = '\t\t\t<h3 class="author">{author} updated:</h3>\n'.format(author=author)
-            author_htm += '\t\t\t<ul class="changes bgimages16">\n'
-            changes_added = []
-            for (css_class, change) in (dictToTuples(e)[0] for e in all_changelog_entries[_date][author]):
-                if change in changes_added: continue
-                write_entry = True
-                changes_added += [change]
-                author_htm += '\t\t\t\t<li class="{css_class}">{change}</li>\n'.format(css_class=css_class, change=change.strip())
-            author_htm += '\t\t\t</ul>\n'
-            if len(changes_added) > 0:
-                entry_htm += author_htm
-        if write_entry:
-            changelog.write(entry_htm)
-
-    with open(os.path.join(targetDir, 'templates', 'footer.html'), 'r') as h:
-        for line in h:
-            changelog.write(line)
-
-
-with open(changelog_cache, 'w') as f:
-    cache_head = 'DO NOT EDIT THIS FILE BY HAND!  AUTOMATICALLY GENERATED BY ss13_genchangelog.py.'
-    yaml.dump_all([cache_head, all_changelog_entries], f, default_flow_style=False)
-
-if len(del_after):
-    print('Cleaning up...')
-    for fileName in del_after:
-        if os.path.isfile(fileName):
-            print(' Deleting {0} (delete-after set)...'.format(fileName))
-            os.remove(fileName)
-
-if errors:
-    sys.exit(1)
diff --git a/tools/pull_request_hooks/autoChangelog.js b/tools/pull_request_hooks/autoChangelog.js
new file mode 100644
index 00000000000..bc24481f182
--- /dev/null
+++ b/tools/pull_request_hooks/autoChangelog.js
@@ -0,0 +1,42 @@
+import { parseChangelog } from "./changelogParser.js";
+
+const safeYml = (string) =>
+	string.replace(/\\/g, "\\\\").replace(/"/g, '\\"').replace(/\n/g, "\\n");
+
+export function changelogToYml(changelog, login) {
+	const author = changelog.author || login;
+	const ymlLines = [];
+
+	ymlLines.push(`author: "${safeYml(author)}"`);
+	ymlLines.push(`delete-after: True`);
+	ymlLines.push(`changes:`);
+
+	for (const change of changelog.changes) {
+		ymlLines.push(
+			`  - ${change.type.changelogKey}: "${safeYml(change.description)}"`
+		);
+	}
+
+	return ymlLines.join("\n");
+}
+
+export async function processAutoChangelog({ github, context }) {
+	const changelog = parseChangelog(context.payload.pull_request.body);
+	if (!changelog || changelog.changes.length === 0) {
+		console.log("no changelog found");
+		return;
+	}
+
+	const yml = changelogToYml(
+		changelog,
+		context.payload.pull_request.user.login
+	);
+
+	github.rest.repos.createOrUpdateFileContents({
+		owner: context.repo.owner,
+		repo: context.repo.repo,
+		path: `html/changelogs/AutoChangeLog-pr-${context.payload.pull_request.number}.yml`,
+		message: `Automatic changelog for PR #${context.payload.pull_request.number} [ci skip]`,
+		content: Buffer.from(yml).toString("base64"),
+	});
+}
diff --git a/tools/pull_request_hooks/autoChangelog.test.js b/tools/pull_request_hooks/autoChangelog.test.js
new file mode 100644
index 00000000000..e7c29c384a9
--- /dev/null
+++ b/tools/pull_request_hooks/autoChangelog.test.js
@@ -0,0 +1,21 @@
+import assert from "node:assert/strict";
+import { changelogToYml } from "./autoChangelog.js";
+import { parseChangelog } from "./changelogParser.js";
+
+assert.equal(
+	changelogToYml(
+		parseChangelog(`
+			My cool PR!
+			:cl: DenverCoder9
+			add: Adds new stuff
+			add: Adds more stuff
+			/:cl:
+		`)
+	),
+
+	`author: "DenverCoder9"
+delete-after: True
+changes:
+  - rscadd: "Adds new stuff"
+  - rscadd: "Adds more stuff"`
+);
diff --git a/tools/pull_request_hooks/changelogConfig.js b/tools/pull_request_hooks/changelogConfig.js
new file mode 100644
index 00000000000..46b28f8d731
--- /dev/null
+++ b/tools/pull_request_hooks/changelogConfig.js
@@ -0,0 +1,120 @@
+/**
+ * A map of changelog phrases to meta-information.
+ *
+ * The first entry in the list is used in the changelog YML file as the key when
+ * used, but other than that all entries are equivalent.
+ *
+ * placeholders - The default messages, if the changelog has this then we pretend it
+ * doesn't exist.
+ */
+export const CHANGELOG_ENTRIES = [
+	[
+		["rscadd", "add", "adds"],
+		{
+			placeholders: [
+				"Added new mechanics or gameplay changes",
+				"Added more things",
+			],
+		},
+	],
+
+	[
+		["bugfix", "fix", "fixes"],
+		{
+			placeholders: ["fixed a few things"],
+		},
+	],
+
+	[
+		["rscdel", "del", "dels"],
+		{
+			placeholders: ["Removed old things"],
+		},
+	],
+
+	[
+		["qol", "tweak"],
+		{
+			placeholders: ["made something easier to use"],
+		},
+	],
+
+    [
+        ["maptweak", "remap"],
+        {
+            placeholders: ["tweaked something in a map"],
+        },
+    ],
+
+	[
+		["sound"],
+		{
+			placeholders: ["added/modified/removed audio or sound effects"],
+		},
+	],
+
+	[
+		["image"],
+		{
+			placeholders: ["added/modified/removed some icons or images"],
+		},
+	],
+
+	[
+		["spellcheck", "typo"],
+		{
+			placeholders: ["fixed a few typos"],
+		},
+	],
+
+	[
+		["balance"],
+		{
+			placeholders: ["rebalanced something"],
+		},
+	],
+
+	[
+		["code_imp", "code"],
+		{
+			placeholders: ["changed some code"],
+		},
+	],
+
+	[
+		["refactor"],
+		{
+			placeholders: ["refactored some code"],
+		},
+	],
+
+	[
+		["config"],
+		{
+			placeholders: ["changed some config setting"],
+		},
+	],
+
+	[
+		["admin"],
+		{
+			placeholders: ["messed with admin stuff"],
+		},
+	],
+
+	[
+		["server"],
+		{
+			placeholders: ["something server ops should know"],
+		},
+	],
+];
+
+// Valid changelog openers
+export const CHANGELOG_OPEN_TAGS = [":cl:", "??"];
+
+// Valid changelog closers
+export const CHANGELOG_CLOSE_TAGS = ["/:cl:", "/ :cl:", ":/cl:", "/??", "/ ??"];
+
+// Placeholder value for an author
+export const CHANGELOG_AUTHOR_PLACEHOLDER_NAME = "optional name here";
diff --git a/tools/pull_request_hooks/changelogParser.js b/tools/pull_request_hooks/changelogParser.js
new file mode 100644
index 00000000000..76e9e216c03
--- /dev/null
+++ b/tools/pull_request_hooks/changelogParser.js
@@ -0,0 +1,80 @@
+import * as changelogConfig from "./changelogConfig.js";
+
+const REGEX_CHANGELOG_LINE = /^(\w+): (.+)$/;
+
+const CHANGELOG_KEYS_TO_ENTRY = {};
+for (const [types, entry] of changelogConfig.CHANGELOG_ENTRIES) {
+	const entryWithChangelogKey = {
+		...entry,
+		changelogKey: types[0],
+	};
+
+	for (const type of types) {
+		CHANGELOG_KEYS_TO_ENTRY[type] = entryWithChangelogKey;
+	}
+}
+
+function parseChangelogBody(lines, openTag) {
+	const [changelogOpening] = lines.splice(0, 1);
+
+	const author =
+		changelogOpening.substring(openTag.length).trim() || undefined;
+
+	const changelog = {
+		author,
+		changes: [],
+	};
+
+	for (const line of lines) {
+		if (line.trim().length === 0) {
+			continue;
+		}
+
+		for (const closeTag of changelogConfig.CHANGELOG_CLOSE_TAGS) {
+			if (line.startsWith(closeTag)) {
+				return changelog;
+			}
+		}
+
+		const match = line.match(REGEX_CHANGELOG_LINE);
+		if (match) {
+			const [_, type, description] = match;
+
+			const entry = CHANGELOG_KEYS_TO_ENTRY[type] || CHANGELOG_KEYS_TO_ENTRY["rscadd"];
+
+			if (entry.placeholders.includes(description)) {
+				continue;
+			}
+
+			if (entry) {
+				changelog.changes.push({
+					type: entry,
+					description,
+				});
+			}
+		} else {
+			const lastChange = changelog.changes[changelog.changes.length - 1];
+			if (lastChange) {
+				lastChange.description += `\n${line}`;
+			}
+		}
+	}
+
+	return changelog;
+}
+
+export function parseChangelog(text) {
+	const lines = text.split("\n").map((line) => line.trim());
+
+	for (let index = 0; index < lines.length; index++) {
+		const line = lines[index];
+
+		for (const openTag of changelogConfig.CHANGELOG_OPEN_TAGS) {
+			if (line.startsWith(openTag)) {
+				return parseChangelogBody(lines.slice(index), openTag);
+			}
+		}
+	}
+
+	return undefined;
+}
diff --git a/tools/pull_request_hooks/changelogParser.test.js b/tools/pull_request_hooks/changelogParser.test.js
new file mode 100644
index 00000000000..bcf044ee36d
--- /dev/null
+++ b/tools/pull_request_hooks/changelogParser.test.js
@@ -0,0 +1,72 @@
+import { strict as assert } from "node:assert";
+import { parseChangelog } from "./changelogParser.js";
+
+// Basic test
+const basicChangelog = parseChangelog(`
+	My cool PR!
+	:cl: DenverCoder9
+	add: Adds new stuff
+	/:cl:
+`);
+
+assert.equal(basicChangelog.author, "DenverCoder9");
+assert.equal(basicChangelog.changes.length, 1);
+assert.equal(basicChangelog.changes[0].type.changelogKey, "rscadd");
+assert.equal(basicChangelog.changes[0].description, "Adds new stuff");
+
+// Multi-line test
+const multiLineChangelog = parseChangelog(`
+	My cool PR!
+	:cl:
+	add: Adds new stuff
+	to the game
+	/:cl:
+`);
+
+assert.equal(multiLineChangelog.author, undefined);
+assert.equal(multiLineChangelog.changes.length, 1);
+assert.equal(multiLineChangelog.changes[0].type.changelogKey, "rscadd");
+assert.equal(
+	multiLineChangelog.changes[0].description,
+	"Adds new stuff\nto the game"
+);
+
+// Placeholders
+const placeholderChangelog = parseChangelog(`
+	My cool PR!
+	:cl:
+	add: Added new mechanics or gameplay changes
+	/:cl:
+`);
+
+assert.equal(placeholderChangelog.changes.length, 0);
+
+// No changelog
+const noChangelog = parseChangelog(`
+	My cool PR!
+`);
+
+assert.equal(noChangelog, undefined);
+
+// No /:cl:
+
+const noCloseChangelog = parseChangelog(`
+	My cool PR!
+	:cl:
+	add: Adds new stuff
+`);
+
+assert.equal(noCloseChangelog.changes.length, 1);
+assert.equal(noCloseChangelog.changes[0].type.changelogKey, "rscadd");
+assert.equal(noCloseChangelog.changes[0].description, "Adds new stuff");
+
+// :cl: with arbitrary text
+
+const arbitraryTextChangelog = parseChangelog(`
+	My cool PR!
+	:cl:
+	Adds new stuff
+	/:cl:
+`);
+
+assert.equal(arbitraryTextChangelog.changes.length, 0);
diff --git a/tools/pull_request_hooks/package.json b/tools/pull_request_hooks/package.json
new file mode 100644
index 00000000000..bedb411a912
--- /dev/null
+++ b/tools/pull_request_hooks/package.json
@@ -0,0 +1,3 @@
+{
+	"type": "module"
+}
diff --git a/tools/pull_request_hooks/removeGuideComments.js b/tools/pull_request_hooks/removeGuideComments.js
new file mode 100644
index 00000000000..c7cdcd1ca67
--- /dev/null
+++ b/tools/pull_request_hooks/removeGuideComments.js
@@ -0,0 +1,41 @@
+import fs from "fs";
+
+const REGEX_COMMENT = /<!--.+?-->/g;
+
+// Make sure we only remove default comments
+const comments = [];
+
+for (const match of fs
+  .readFileSync(".github/PULL_REQUEST_TEMPLATE.md", { encoding: "utf8" })
+  .matchAll(REGEX_COMMENT)) {
+  comments.push(match[0]);
+}
+
+function escapeRegex(string) {
+  return string.replace(/[-\/\\^$*+?.()|[\]{}]/g, "\\$&");
+}
+
+export async function removeGuideComments({ github, context }) {
+  let newBody = context.payload.pull_request.body;
+
+  if (!newBody) {
+	console.log("PR body is empty, skipping...");
+	return;
+  }
+
+  for (const comment of comments) {
+    newBody = newBody.replace(
+      new RegExp(`^\\s*${escapeRegex(comment)}\\s*`, "gm"),
+      "\n"
+    );
+  }
+
+  if (newBody !== context.payload.pull_request.body) {
+    await github.rest.pulls.update({
+      pull_number: context.payload.pull_request.number,
+      repo: context.repo.repo,
+      owner: context.repo.owner,
+      body: newBody,
+    });
+  }
+}
diff --git a/tools/pull_request_hooks/rerunFlakyTests.js b/tools/pull_request_hooks/rerunFlakyTests.js
new file mode 100644
index 00000000000..6625d9caa8c
--- /dev/null
+++ b/tools/pull_request_hooks/rerunFlakyTests.js
@@ -0,0 +1,293 @@
+const LABEL = "🤖 Flaky Test Report";
+const TITLE_BOT_HEADER = "title: ";
+
+// Only check jobs that start with these.
+// Helps make sure we don't restart something like screenshot tests or linters, which are not known to be flaky.
+const CONSIDERED_JOBS = [
+	"Integration Tests",
+];
+
+async function getFailedJobsForRun(github, context, workflowRunId, runAttempt) {
+	const {
+		data: { jobs },
+	} = await github.rest.actions.listJobsForWorkflowRunAttempt({
+		owner: context.repo.owner,
+		repo: context.repo.repo,
+		run_id: workflowRunId,
+		attempt_number: runAttempt,
+	});
+
+	return jobs
+		.filter((job) => job.conclusion === "failure")
+		.filter((job) =>
+			CONSIDERED_JOBS.some((title) => job.name.startsWith(title))
+		);
+}
+
+export async function rerunFlakyTests({ github, context }) {
+	const failingJobs = await getFailedJobsForRun(
+		github,
+		context,
+		context.payload.workflow_run.id,
+		context.payload.workflow_run.run_attempt
+	);
+
+	if (failingJobs.length > 1) {
+		console.log("Multiple jobs failing. PROBABLY not flaky, not rerunning.");
+		return;
+	}
+
+	if (failingJobs.length === 0) {
+		throw new Error(
+			"rerunFlakyTests should not have run on a run with no failing jobs"
+		);
+	}
+
+	github.rest.actions.reRunWorkflowFailedJobs({
+		owner: context.repo.owner,
+		repo: context.repo.repo,
+		run_id: context.payload.workflow_run.id,
+	});
+}
+
+// Tries its best to extract a useful error title and message for the given log
+export function extractDetails(log) {
+	// Strip off timestamp
+	const lines = log.split(/^[0-9.:T\-]*?Z /gm);
+
+	const failureRegex = /^\t?FAILURE #(?<number>[0-9]+): (?<headline>.+)/;
+	const groupRegex = /^##\[group\](?<group>.+)/;
+
+	const failures = [];
+	let lastGroup = "root";
+	let loggingFailure;
+
+	const newFailure = (failureMatch) => {
+		const { headline } = failureMatch.groups;
+
+		loggingFailure = {
+			headline,
+			group: lastGroup.replace("/datum/unit_test/", ""),
+			details: [],
+		};
+	};
+
+	for (const line of lines) {
+		const groupMatch = line.match(groupRegex);
+		if (groupMatch) {
+			lastGroup = groupMatch.groups.group.trim();
+			continue;
+		}
+
+		const failureMatch = line.match(failureRegex);
+
+		if (loggingFailure === undefined) {
+			if (!failureMatch) {
+				continue;
+			}
+
+			newFailure(failureMatch);
+		} else if (failureMatch || line.startsWith("##")) {
+			failures.push(loggingFailure);
+			loggingFailure = undefined;
+
+			if (failureMatch) {
+				newFailure(failureMatch);
+			}
+		} else {
+			loggingFailure.details.push(line.trim());
+		}
+	}
+
+	// We had no logged failures, there's not really anything we can do here
+	if (failures.length === 0) {
+		return {
+			title: "Flaky test failure with no obvious source",
+			failures,
+		};
+	}
+
+	// We *could* create multiple failures for multiple groups.
+	// This would be important if we had multiple flaky tests at the same time.
+	// I'm choosing not to because it complicates this logic a bit, has the ability to go terribly wrong,
+	// and also because there's something funny to me about that increasing the urgency of fixing
+	// flaky tests. If it becomes a serious issue though, I would not mind this being fixed.
+	const uniqueGroups = new Set(failures.map((failure) => failure.group));
+
+	if (uniqueGroups.size > 1) {
+		return {
+			title: `Multiple flaky test failures in ${Array.from(uniqueGroups)
+				.sort()
+				.join(", ")}`,
+			failures,
+		};
+	}
+
+	const failGroup = failures[0].group;
+
+	if (failures.length > 1) {
+		return {
+			title: `Multiple errors in flaky test ${failGroup}`,
+			failures,
+		};
+	}
+
+	const failure = failures[0];
+
+	// Common patterns where we can always get a detailed title
+	const runtimeMatch = failure.headline.match(/Runtime in .+?: (?<error>.+)/);
+	if (runtimeMatch) {
+		const runtime = runtimeMatch.groups.error.trim();
+
+		const invalidTimerMatch = runtime.match(/^Invalid timer:.+object:(?<object>[^[]+).*delegate:(?<proc>.+?), source:/);
+		if (invalidTimerMatch) {
+			return {
+				title: `Flaky test ${failGroup}: Invalid timer: ${invalidTimerMatch.groups.proc.trim()} on ${invalidTimerMatch.groups.object.trim()}`,
+				failures,
+			};
+		}
+
+		return {
+			title: `Flaky test ${failGroup}: ${runtime}`,
+			failures,
+		};
+	}
+
+	const hardDelMatch = failure.headline.match(/^(?<object>\/[\w/]+) hard deleted .* times out of a total del count of/);
+	if (hardDelMatch) {
+		return {
+			title: `Flaky hard delete: ${hardDelMatch.groups.object}`,
+			failures,
+		};
+	}
+
+	// Try to normalize the title and remove anything that might be variable
+	const normalizedError = failure.headline.replace(/\s*at .+?:[0-9]+.*/g, ""); // "<message> at code.dm:123"
+
+	return {
+		title: `Flaky test ${failGroup}: ${normalizedError}`,
+		failures,
+	};
+}
+
+async function getExistingIssueId(graphql, context, title) {
+	// Hope you never have more than 100 of these open!
+	const {
+		repository: {
+			issues: { nodes: openFlakyTestIssues },
+		},
+	} = await graphql(
+		`
+			query ($owner: String!, $repo: String!, $label: String!) {
+				repository(owner: $owner, name: $repo) {
+					issues(
+						labels: [$label]
+						first: 100
+						orderBy: { field: CREATED_AT, direction: DESC }
+						states: [OPEN]
+					) {
+						nodes {
+							number
+							title
+							body
+						}
+					}
+				}
+			}
+		`,
+		{
+			owner: context.repo.owner,
+			repo: context.repo.repo,
+			label: LABEL,
+		}
+	);
+
+	const exactTitle = openFlakyTestIssues.find((issue) => issue.title === title);
+	if (exactTitle !== undefined) {
+		return exactTitle.number;
+	}
+
+	const foundInBody = openFlakyTestIssues.find((issue) =>
+		issue.body.includes(`<!-- ${TITLE_BOT_HEADER}${exactTitle} -->`)
+	);
+	if (foundInBody !== undefined) {
+		return foundInBody.number;
+	}
+
+	return undefined;
+}
+
+function createBody({ title, failures }, runUrl) {
+	return `
+	<!-- This issue can be renamed, but do not change the next comment! -->
+	<!-- title: ${title} -->
+
+	Flaky tests were detected in [this test run](${runUrl}). This means that there was a failure that was cleared when the tests were simply restarted.
+
+	Failures:
+	\`\`\`
+	${failures
+		.map(
+			(failure) =>
+				`${failure.group}: ${failure.headline}\n\t${failure.details.join("\n")}`
+		)
+		.join("\n")}
+	\`\`\`
+	`.replace(/^\s*/gm, "");
+}
+
+export async function reportFlakyTests({ github, context }) {
+	const failedJobsFromLastRun = await getFailedJobsForRun(
+		github,
+		context,
+		context.payload.workflow_run.id,
+		context.payload.workflow_run.run_attempt - 1
+	);
+
+	// This could one day be relaxed if we face serious enough flaky test problems, so we're going to loop anyway
+	if (failedJobsFromLastRun.length !== 1) {
+		console.log(
+			"Multiple jobs failing after retry, assuming maintainer rerun."
+		);
+
+		return;
+	}
+
+	for (const job of failedJobsFromLastRun) {
+		const { data: log } =
+			await github.rest.actions.downloadJobLogsForWorkflowRun({
+				owner: context.repo.owner,
+				repo: context.repo.repo,
+				job_id: job.id,
+			});
+
+		const details = extractDetails(log);
+
+		const existingIssueId = await getExistingIssueId(
+			github.graphql,
+			context,
+			details.title
+		);
+
+		if (existingIssueId !== undefined) {
+			// Maybe in the future, if it's helpful, update the existing issue with new links
+			console.log(`Existing issue found: #${existingIssueId}`);
+			return;
+		}
+
+		await github.rest.issues.create({
+			owner: context.repo.owner,
+			repo: context.repo.repo,
+			title: details.title,
+			labels: [LABEL],
+			body: createBody(
+				details,
+				`https://github.com/${context.repo.owner}/${
+					context.repo.repo
+				}/actions/runs/${context.payload.workflow_run.id}/attempts/${
+					context.payload.workflow_run.run_attempt - 1
+				}`
+			),
+		});
+	}
+}
diff --git a/tools/pull_request_hooks/rerunFlakyTests.test.js b/tools/pull_request_hooks/rerunFlakyTests.test.js
new file mode 100644
index 00000000000..6a3300ff59b
--- /dev/null
+++ b/tools/pull_request_hooks/rerunFlakyTests.test.js
@@ -0,0 +1,45 @@
+import { strict as assert } from "node:assert";
+import fs from "node:fs";
+import { extractDetails } from "./rerunFlakyTests.js";
+
+function extractDetailsFromPayload(filename) {
+	return extractDetails(
+		fs.readFileSync(`tests/flakyTestPayloads/${filename}.txt`, {
+			encoding: "utf8",
+		})
+	);
+}
+
+const chatClient = extractDetailsFromPayload("chat_client");
+assert.equal(
+	chatClient.title,
+	"Flaky hard delete: /datum/computer_file/program/chatclient"
+);
+assert.equal(chatClient.failures.length, 1);
+
+const monkeyBusiness = extractDetailsFromPayload("monkey_business");
+assert.equal(
+	monkeyBusiness.title,
+	"Flaky test monkey_business: Cannot execute null.resolve()."
+);
+assert.equal(monkeyBusiness.failures.length, 1);
+
+const shapeshift = extractDetailsFromPayload("shapeshift");
+assert.equal(
+	shapeshift.title,
+	"Multiple errors in flaky test shapeshift_spell"
+);
+assert.equal(shapeshift.failures.length, 16);
+
+const multipleFailures = extractDetailsFromPayload("multiple_failures");
+assert.equal(
+	multipleFailures.title,
+	"Multiple flaky test failures in more_shapeshift_spell, shapeshift_spell"
+);
+assert.equal(multipleFailures.failures.length, 2);
+
+const invalidTimer = extractDetailsFromPayload("invalid_timer");
+assert.equal(
+	invalidTimer.title,
+	"Flaky test monkey_business: Invalid timer: /datum/looping_sound/proc/start_sound_loop() on /datum/looping_sound/showering"
+);