diff --git a/scripts/sarif-download-projects.py b/scripts/sarif-download-projects.py
new file mode 100755
index 0000000..9252a35
--- /dev/null
+++ b/scripts/sarif-download-projects.py
@@ -0,0 +1,45 @@
+#!/usr/bin/env python3
+""" This is part 1 of 2 hardcoded utility scripts.  This one downloads the first
+    1000 project pages from lgtm.com and saves the `projects` information in
+    ~/local/sarif/projects.pickle for use by `sarif-download-sarif.py`
+"""
+import concurrent.futures
+import pathlib
+import pickle
+import requests
+import sys
+
+LGTM_URL = "https://lgtm.com/"
+SESSION = requests.Session()
+
+OUTPUT_DIRECTORY = pathlib.Path(__file__).parent
+OUTPUT_DIRECTORY = pathlib.Path.home() / "local/sarif"
+PROJECT_FILE = OUTPUT_DIRECTORY / "projects.pickle"
+
+if PROJECT_FILE.exists():
+    sys.stderr.write("error: output file %s exists\n" % PROJECT_FILE)
+    sys.exit(1)
+
+OUTPUT_DIRECTORY.mkdir(mode=0o755, parents=True, exist_ok=True)
+
+projects = {}
+page = 1
+current_projects_url = "%sapi/v1.0/projects/" % LGTM_URL
+while page < 1000:
+	print("Fetching projects page %d..." % page)
+	page += 1
+	response = SESSION.get(current_projects_url)
+	response.raise_for_status()
+	response_data = response.json()
+	for item in response_data["data"]:
+		projects[item["id"]] = item
+	if "nextPageUrl" in response_data:
+		current_projects_url = response_data["nextPageUrl"]
+	else:
+		break
+
+# Save them
+with open(PROJECT_FILE, 'wb') as outfile:
+    pickle.dump(projects, outfile)
+
+print("All projects fetched, saved to %s" % PROJECT_FILE)
diff --git a/scripts/sarif-download-sarif.py b/scripts/sarif-download-sarif.py
new file mode 100755
index 0000000..70101ef
--- /dev/null
+++ b/scripts/sarif-download-sarif.py
@@ -0,0 +1,73 @@
+#!/usr/bin/env python3
+""" This is part 2 of 2 hardcoded utility scripts.  This one downloads the sarif
+    files for the `projects` collected by `sarif-download-projects.py` to
+    subdirectories of ~/local/sarif/projects.pickle.
+    
+    Already downloaded files will not be tried again, so if this script fails it
+    can just be rerun.
+"""
+
+import concurrent.futures
+import pathlib
+import pickle
+import requests
+import sys
+
+LGTM_URL = "https://lgtm.com/"
+SESSION = requests.Session()
+
+# OUTPUT_DIRECTORY = pathlib.Path(__file__).parent
+OUTPUT_DIRECTORY = pathlib.Path.home() / "local/sarif"
+PROJECT_FILE = OUTPUT_DIRECTORY / "projects.pickle"
+
+if not PROJECT_FILE.exists():
+    sys.stderr.write("error: missing input file %s\n" % PROJECT_FILE)
+    sys.exit(1)
+
+OUTPUT_DIRECTORY.mkdir(mode=0o755, parents=True, exist_ok=True)
+
+with open(PROJECT_FILE, "rb") as infile:
+    projects = pickle.load(infile)
+
+thread_pool = concurrent.futures.ThreadPoolExecutor(25)
+futures = {}
+any_failed = []
+def process(index, pair):
+    try:
+        project_key, project = pair
+        output_path = OUTPUT_DIRECTORY / project["url-identifier"] / "results.sarif"
+        if output_path.exists():
+            print("Already fetched %d/%d (%s)" % 
+                  (index + 1, len(projects), project["url-identifier"]))
+            return
+        else:
+            print("Processing project %d/%d (%s)..." %                   
+                  (index + 1, len(projects), project["url-identifier"]))
+
+        # Get latest analysis information.
+        analysis_summary_url = "%sapi/v1.0/analyses/%d/commits/latest" % (LGTM_URL, project_key)
+        response = SESSION.get(analysis_summary_url)
+        response.raise_for_status()
+        analysis_summary = response.json()
+        analysis_id = analysis_summary["id"]
+
+        # Get SARIF export.
+        sarif_url = "%sapi/v1.0/analyses/%s/alerts" % (LGTM_URL, analysis_id)
+        response = SESSION.get(sarif_url)
+        response.raise_for_status()
+        # And save it 
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+        output_path.write_text(response.text)
+    except:
+        any_failed.append( (index, pair) )
+
+for index, pair in enumerate(projects.items()):
+    try:
+        futures[pair[0]] = thread_pool.submit(process, index, pair)
+    except RuntimeError:
+        pass
+thread_pool.shutdown()
+
+for index, pair in any_failed:
+    print("Processing failed for %d, %s:" % (index, pair))
+    print("Re-run to try those again")