VS-1779 clean up parquet files immediatement by gbggrant · Pull Request #9338 · broadinstitute/gatk

gbggrant · 2026-02-26T17:39:30Z

This PR adds functionality to (by default) delete all of the parquet files generated during import.

Passing run here

…ParquetFilesImmediatement

Copilot

Pull request overview

Adds an option to automatically clean up Parquet artifacts produced during the VariantStore import workflow, aiming to reduce storage usage after a successful load.

Changes:

Introduces delete_parquet_files_after_loading (default true) to control Parquet cleanup post-load.
Adds a DeleteParquetFiles WDL task and wires it into GvsImportGenomes.
Updates Dockstore workflow branch filters.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 6 comments.

File	Description
scripts/variantstore/wdl/GvsImportGenomes.wdl	Adds a new workflow input and a cleanup task intended to delete staged Parquet files after loading.
scripts/variantstore/wdl/GvsBulkIngestGenomes.wdl	Adds a stray comment line near imports.
.dockstore.yml	Adds a branch filter entry for the new branch name.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

scripts/variantstore/wdl/GvsImportGenomes.wdl

scripts/variantstore/wdl/GvsBulkIngestGenomes.wdl

.dockstore.yml

scripts/variantstore/wdl/GvsImportGenomes.wdl

Copilot

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 8 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

scripts/variantstore/wdl/GvsImportGenomes.wdl

Copilot · 2026-02-26T19:14:02Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+  runtime {
+    docker: cloud_sdk_docker
+    memory: "3 GB"
+    disks: "local-disk 500 HDD"


The task allocates 500 GB of disk space ("local-disk 500 HDD"), which seems excessive for a task that only lists and deletes GCS objects. The script only creates small text files locally (parquet_dirs.txt). Consider reducing the disk allocation to something more reasonable like 10-20 GB, consistent with similar tasks like ConfigureParquetLifecycle which uses 10 GB.

Suggested change

disks: "local-disk 500 HDD"

disks: "local-disk 10 HDD"

scripts/variantstore/wdl/GvsImportGenomes.wdl

.dockstore.yml

Copilot · 2026-02-26T19:14:03Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+
+    # List the contents of the vet and ref_ranges directories for deletion later
+    echo "Listing directories under ${OUTPUT_GCS_DIR}/vet/ and ${OUTPUT_GCS_DIR}/ref_ranges/ for deletion..."
+    gcloud storage ls  ~{"--billing-project " + billing_project_id} \


The command uses double spacing between "ls" and the billing_project_id flag. While this may work, it's inconsistent with line 1256 which has single spacing. Consider using consistent spacing for better code style.

Suggested change

gcloud storage ls ~{"--billing-project " + billing_project_id} \

gcloud storage ls ~{"--billing-project " + billing_project_id} \

scripts/variantstore/wdl/GvsImportGenomes.wdl

Copilot · 2026-02-26T19:14:03Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+    gcloud storage ls  ~{"--billing-project " + billing_project_id} \
+    "${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt || true
+
+    # Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein
+    echo "Deleting objects in vet and ref_ranges directories..."
+    while IFS= read -r gcs_path; do
+      if [ -n "$gcs_path" ]; then
+        echo "Deleting objects in: $gcs_path"
+        gcloud storage rm ~{"--billing-project " + billing_project_id} \
+          "${gcs_path}**" --recursive || echo "Warning: Failed to delete $gcs_path"


The script lists directories using "gcloud storage ls" which outputs directory paths, but then attempts to delete "${gcs_path}" recursively. Since gcs_path will contain directory paths ending with "/", the resulting pattern "${gcs_path}" may not match the intended files. Additionally, the error handling using "|| true" on line 1249 will suppress any listing errors, potentially leading to silent failures if the directories don't exist. Consider handling the case where directories might not exist more explicitly.

Suggested change

gcloud storage ls ~{"--billing-project " + billing_project_id} \

"${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt || true

# Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein

echo "Deleting objects in vet and ref_ranges directories..."

while IFS= read -r gcs_path; do

if [ -n "$gcs_path" ]; then

echo "Deleting objects in: $gcs_path"

gcloud storage rm ~{"--billing-project " + billing_project_id} \

"${gcs_path}**" --recursive || echo "Warning: Failed to delete $gcs_path"

if ! gcloud storage ls ~{"\"--billing-project " + billing_project_id + "\"" } \

"${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt; then

echo "Warning: Failed to list parquet directories under ${OUTPUT_GCS_DIR}/vet/ and ${OUTPUT_GCS_DIR}/ref_ranges/. They may not exist."

# Ensure parquet_dirs.txt exists but is empty so the deletion loop is a no-op

: > parquet_dirs.txt

fi

# Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein

echo "Deleting objects in vet and ref_ranges directories..."

while IFS= read -r gcs_path; do

if [ -n "$gcs_path" ]; then

# Normalize to remove any trailing slash before appending '/**'

gcs_prefix="${gcs_path%/}"

echo "Deleting objects in: $gcs_prefix"

gcloud storage rm ~{"\"--billing-project " + billing_project_id + "\"" } \

"${gcs_prefix}/**" --recursive || echo "Warning: Failed to delete $gcs_path"

Copilot

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 5 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

scripts/variantstore/wdl/GvsImportGenomes.wdl

Copilot · 2026-02-26T19:36:13Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+    # Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein
+    echo "Deleting objects in vet and ref_ranges directories..."
+    while IFS= read -r gcs_path; do
+      if [ -n "$gcs_path" ]; then
+        echo "Deleting objects in: $gcs_path"
+        gcloud storage rm ~{"--billing-project " + billing_project_id} "$gcs_path" --recursive
+      fi
+    done < parquet_dirs.txt
+
+    echo "✓ Completed deletion of vet and ref_ranges parquet files"
+


The deletion command at line 1256 will fail if gcloud storage rm encounters an error (e.g., permission denied, network issues), which is the desired behavior. However, consider adding error handling or at least a check that verifies some files were actually found and deleted. Currently, if the ls command on line 1248 returns no results (even due to an access error suppressed by || true), the task will silently succeed without deleting anything. Consider adding validation that parquet_dirs.txt is not empty before proceeding with deletion, or logging a warning if no files are found.

Suggested change

# Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein

echo "Deleting objects in vet and ref_ranges directories..."

while IFS= read -r gcs_path; do

if [ -n "$gcs_path" ]; then

echo "Deleting objects in: $gcs_path"

gcloud storage rm ~{"--billing-project " + billing_project_id} "$gcs_path" --recursive

fi

done < parquet_dirs.txt

echo "✓ Completed deletion of vet and ref_ranges parquet files"

# Validate that we found some directories before attempting deletion

if [ ! -s parquet_dirs.txt ]; then

echo "WARNING: No vet or ref_ranges directories found to delete. This may indicate that no parquet files were generated or there was an access issue listing the bucket." >&2

else

# Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein

echo "Deleting objects in vet and ref_ranges directories..."

deleted_any=false

while IFS= read -r gcs_path; do

if [ -n "$gcs_path" ]; then

deleted_any=true

echo "Deleting objects in: $gcs_path"

gcloud storage rm ~{"--billing-project " + billing_project_id} "$gcs_path" --recursive

fi

done < parquet_dirs.txt

if [ "$deleted_any" = true ]; then

echo "✓ Completed deletion of vet and ref_ranges parquet files"

else

echo "WARNING: parquet_dirs.txt contained no valid GCS paths; no objects were deleted." >&2

fi

fi

scripts/variantstore/wdl/GvsImportGenomes.wdl

…ParquetFilesImmediatement

gatk-bot · 2026-03-04T21:49:02Z

Github actions tests reported job failures from actions build 22689373565
Failures in the following jobs:

Test Type	JDK	Job ID	Logs
conda	17.0.6+10	22689373565.3	logs

mcovarr · 2026-03-05T15:06:49Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+    # List the contents of the vet and ref_ranges directories for subsequent deletion in the loop below
+    echo "Listing directories under ${OUTPUT_GCS_DIR}/vet/ and ${OUTPUT_GCS_DIR}/ref_ranges/ for deletion..."
+    gcloud storage ls ~{"--billing-project " + billing_project_id} \
+    "${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt


Why is this ls necessary? couldn't we just

gcloud storage rm ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}"'/*' --recursive

Beyond the simplification, this would also have the advantage of clearing out the sample_chromosome_ploidy data that the code is currently not dealing with, plus the header data that may appear someday.

Likely because mass deletion commands in gcp regularly short circuit and throw errors when trying to delete too many individual items. When I tried to manually clean up after running the 175k exome callset, attempting to delete the entire root of the directories failed repeatedly. Even trying to delete just the vet or ref data as a whole failed. Only removing each subdirectory was able to reliably work, sadly

As discussed in standup, I am following the suggested pattern that @koncheto-broad had suggested in the ticket. Happy to simplify if that is acceptable to all.

I'm surprised there were issues with a recursive delete of 100Ks of objects as I've successfully deleted millions of objects without issue 🤷

scripts/variantstore/wdl/GvsImportGenomes.wdl

Co-authored-by: Miguel Covarrubias <mcovarr@users.noreply.github.com>

…ParquetFilesImmediatement

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

You can also share your feedback on Copilot code review. Take the survey.

Copilot · 2026-03-09T16:26:43Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+#    # List the contents of the vet and ref_ranges directories for subsequent deletion in the loop below
+#    echo "Listing directories under ${OUTPUT_GCS_DIR}/vet/ and ${OUTPUT_GCS_DIR}/ref_ranges/ for deletion..."
+#    gcloud storage ls ~{"--billing-project " + billing_project_id} \
+#    "${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt
+#
+#    # Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein
+#    echo "Deleting Parquet files..."
+#    while IFS= read -r gcs_path; do
+#      if [ -n "$gcs_path" ]; then
+#        echo "Deleting objects in: $gcs_path"
+#        gcloud storage rm ~{"--billing-project " + billing_project_id} "$gcs_path" --recursive
+#      fi
+#    done < parquet_dirs.txt
+


There is a large block of commented-out shell logic left in the task command. Please remove the dead/commented code (or convert it into the active implementation) to keep the task readable and avoid confusion about the intended deletion strategy.

Suggested change

# # List the contents of the vet and ref_ranges directories for subsequent deletion in the loop below

# echo "Listing directories under ${OUTPUT_GCS_DIR}/vet/ and ${OUTPUT_GCS_DIR}/ref_ranges/ for deletion..."

# gcloud storage ls ~{"--billing-project " + billing_project_id} \

# "${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt

#

# # Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein

# echo "Deleting Parquet files..."

# while IFS= read -r gcs_path; do

# if [ -n "$gcs_path" ]; then

# echo "Deleting objects in: $gcs_path"

# gcloud storage rm ~{"--billing-project " + billing_project_id} "$gcs_path" --recursive

# fi

# done < parquet_dirs.txt

I agree this shouldn't remain commented out; up to you whether it's enabled or not 🙂

Copilot · 2026-03-09T16:26:44Z

.dockstore.yml

         - master
         - ah_var_store
-         - gg_VS-1794_ParquetRemovalStrategy
+         - gg_VS-1779_CleanUpParquetFilesImmediatement


Spelling: branch name gg_VS-1779_CleanUpParquetFilesImmediatement contains a typo (“Immediatement”). If possible, rename/update to “Immediately” to keep branch filters consistent and avoid future confusion.

Suggested change

- gg_VS-1779_CleanUpParquetFilesImmediatement

- gg_VS-1779_CleanUpParquetFilesImmediately

Mais non mon robot, on parle français ici !

Copilot · 2026-03-09T16:26:44Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+    gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"
+
+#    # List the contents of the vet and ref_ranges directories for subsequent deletion in the loop below
+#    echo "Listing directories under ${OUTPUT_GCS_DIR}/vet/ and ${OUTPUT_GCS_DIR}/ref_ranges/ for deletion..."
+#    gcloud storage ls ~{"--billing-project " + billing_project_id} \
+#    "${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt
+#
+#    # Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein
+#    echo "Deleting Parquet files..."
+#    while IFS= read -r gcs_path; do
+#      if [ -n "$gcs_path" ]; then
+#        echo "Deleting objects in: $gcs_path"
+#        gcloud storage rm ~{"--billing-project " + billing_project_id} "$gcs_path" --recursive
+#      fi
+#    done < parquet_dirs.txt
+


DeleteParquetFiles deletes everything under output_gcs_dir via gcloud storage rm --recursive ... "${OUTPUT_GCS_DIR}/", not just the Parquet objects. Since parquet_output_gcs_dir is user-provided, a mis-specified prefix could recursively delete unrelated data. Consider deleting only the discovered .parquet URLs (e.g., pass DiscoverParquetFiles.all_files_list into this task and delete those), or restrict deletion to the known Parquet subpaths (vet/, ref_ranges/, sample_chromosome_ploidy/) instead of the entire prefix.

Suggested change

gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"

# # List the contents of the vet and ref_ranges directories for subsequent deletion in the loop below

# echo "Listing directories under ${OUTPUT_GCS_DIR}/vet/ and ${OUTPUT_GCS_DIR}/ref_ranges/ for deletion..."

# gcloud storage ls ~{"--billing-project " + billing_project_id} \

# "${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt

#

# # Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein

# echo "Deleting Parquet files..."

# while IFS= read -r gcs_path; do

# if [ -n "$gcs_path" ]; then

# echo "Deleting objects in: $gcs_path"

# gcloud storage rm ~{"--billing-project " + billing_project_id} "$gcs_path" --recursive

# fi

# done < parquet_dirs.txt

# Delete only known Parquet subdirectories under the output GCS directory,

# instead of recursively deleting everything under OUTPUT_GCS_DIR.

for SUBDIR in vet ref_ranges sample_chromosome_ploidy; do

TARGET_DIR="${OUTPUT_GCS_DIR}/${SUBDIR}/"

echo "Checking for Parquet directory: ${TARGET_DIR}"

if gcloud storage ls ~{"\"--billing-project " + billing_project_id + "\""} "${TARGET_DIR}" >/dev/null 2>&1; then

echo "Deleting objects under: ${TARGET_DIR}"

gcloud storage rm --recursive ~{"\"--billing-project " + billing_project_id + "\""} "${TARGET_DIR}"

else

echo "No objects found under: ${TARGET_DIR}, skipping."

fi

done

mcovarr · 2026-03-09T16:43:02Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+    # Normalize GCS path by removing any trailing slash
+    OUTPUT_GCS_DIR=$(echo ~{output_gcs_dir} | sed 's/\/$//')
+
+    gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"


To address some of Copilot's specificity concerns:

Suggested change

gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"

gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"'**/*.parquet'

mcovarr · 2026-03-09T16:43:47Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+#    # List the contents of the vet and ref_ranges directories for subsequent deletion in the loop below
+#    echo "Listing directories under ${OUTPUT_GCS_DIR}/vet/ and ${OUTPUT_GCS_DIR}/ref_ranges/ for deletion..."
+#    gcloud storage ls ~{"--billing-project " + billing_project_id} \
+#    "${OUTPUT_GCS_DIR}/vet/" "${OUTPUT_GCS_DIR}/ref_ranges/" > parquet_dirs.txt
+#
+#    # Iterate over all Google Cloud paths in parquet_dirs.txt and delete all objects therein
+#    echo "Deleting Parquet files..."
+#    while IFS= read -r gcs_path; do
+#      if [ -n "$gcs_path" ]; then
+#        echo "Deleting objects in: $gcs_path"
+#        gcloud storage rm ~{"--billing-project " + billing_project_id} "$gcs_path" --recursive
+#      fi
+#    done < parquet_dirs.txt
+


I agree this shouldn't remain commented out; up to you whether it's enabled or not 🙂

mcovarr · 2026-03-09T18:55:53Z

scripts/variantstore/wdl/GvsImportGenomes.wdl

+    OUTPUT_GCS_DIR=$(echo ~{output_gcs_dir} | sed 's/\/$//')
+
+    if [ "~{use_alternate_delete_strategy}" = "false" ]; then
+      gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"**/*.parquet


going to need some single quotes or escapes here else the globbing is likely to be a problem

Suggested change

gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"**/*.parquet

gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"'**/*.parquet'

gbggrant added 6 commits February 25, 2026 13:36

WIP

41853ff

Initial cruft

a59371b

Test before live fire

e15571d

slight update

e215f37

The real show.

8981039

Merge remote-tracking branch 'origin/VS-1736' into gg_VS-1779_CleanUp…

1c9a790

…ParquetFilesImmediatement

gbggrant requested a review from Copilot February 26, 2026 17:39

Copilot started reviewing on behalf of gbggrant February 26, 2026 17:40 View session

Copilot AI reviewed Feb 26, 2026

View reviewed changes

gbggrant added 2 commits February 26, 2026 14:07

Address copilot's code review suggestions.

3471d89

Another thing

d977fcf

gbggrant requested a review from Copilot February 26, 2026 19:09

Copilot started reviewing on behalf of gbggrant February 26, 2026 19:10 View session

Copilot AI reviewed Feb 26, 2026

View reviewed changes

some more

9006292

gbggrant requested a review from Copilot February 26, 2026 19:31

Copilot started reviewing on behalf of gbggrant February 26, 2026 19:32 View session

Copilot AI reviewed Feb 26, 2026

View reviewed changes

gbggrant added 4 commits March 4, 2026 11:30

Merge remote-tracking branch 'origin/VS-1736' into gg_VS-1779_CleanUp…

070fc29

…ParquetFilesImmediatement

Code review suggestions.

255c487

Point to new branch in dockstore.

ad3204d

Last

6a20fe4

gbggrant marked this pull request as ready for review March 4, 2026 21:09

gbggrant requested review from koncheto-broad and mcovarr March 4, 2026 21:09

mcovarr reviewed Mar 5, 2026

View reviewed changes

gbggrant and others added 3 commits March 5, 2026 10:40

Update scripts/variantstore/wdl/GvsImportGenomes.wdl

5c92904

Co-authored-by: Miguel Covarrubias <mcovarr@users.noreply.github.com>

Update scripts/variantstore/wdl/GvsImportGenomes.wdl

f6c7380

Co-authored-by: Miguel Covarrubias <mcovarr@users.noreply.github.com>

Merge remote-tracking branch 'origin/VS-1736' into gg_VS-1779_CleanUp…

a4a0134

…ParquetFilesImmediatement

gbggrant added 2 commits March 9, 2026 09:51

Merge remote-tracking branch 'origin/VS-1736' into gg_VS-1779_CleanUp…

7727358

…ParquetFilesImmediatement

Change to doing a simple rm -r on the whole directory

a3eeb1d

gbggrant requested a review from Copilot March 9, 2026 16:23

Copilot started reviewing on behalf of gbggrant March 9, 2026 16:23 View session

Copilot AI reviewed Mar 9, 2026

View reviewed changes

mcovarr approved these changes Mar 9, 2026

View reviewed changes

gbggrant added 4 commits March 9, 2026 14:34

Allow for alternate deletion strategy

a2e1833

Push up the flag to allow for alternate deletion strategy.

27ed4b9

Push up the flag to allow for alternate deletion strategy.

8f10595

I missed this one.

db066cc

mcovarr reviewed Mar 9, 2026

View reviewed changes

gbggrant added 4 commits March 9, 2026 14:56

MOre

1845747

Last details

a114eb8

One last little thing

e8faf9a

Cleanup

38ad4ce

gbggrant merged commit 71a586d into VS-1736 Mar 10, 2026
7 of 12 checks passed

gbggrant deleted the gg_VS-1779_CleanUpParquetFilesImmediatement branch March 10, 2026 12:15

	gcloud storage ls ~{"--billing-project " + billing_project_id} \
	gcloud storage ls ~{"--billing-project " + billing_project_id} \

	- gg_VS-1779_CleanUpParquetFilesImmediatement
	- gg_VS-1779_CleanUpParquetFilesImmediately

	gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"
	gcloud storage rm --recursive ~{"--billing-project " + billing_project_id} "${OUTPUT_GCS_DIR}/"'*/.parquet'

Conversation

gbggrant commented Feb 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

gatk-bot commented Mar 4, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 9, 2026

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 9, 2026

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 9, 2026

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

gbggrant commented Feb 26, 2026 •

edited

Loading