feat(RELEASE-2479): add collect_index_images.py script by johnbieren · Pull Request #851 · konflux-ci/release-service-utils

johnbieren · 2026-06-24T18:37:20Z

This commit adds a python script to replicate the functionality of the inline bash script in the collect-index-images managed task in the catalog repo. The task will be updated to use this python module instead.

Assisted-By: Cursor

codecov-commenter · 2026-06-24T18:41:06Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 95.70%. Comparing base (a0c68a7) to head (8896c96).
⚠️ Report is 6 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #851      +/-   ##
==========================================
+ Coverage   95.51%   95.70%   +0.18%     
==========================================
  Files          69       72       +3     
  Lines        6893     7192     +299     
==========================================
+ Hits         6584     6883     +299     
  Misses        309      309

Flag	Coverage Δ
unit-tests	`95.70% <100.00%> (+0.18%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
scripts/python/helpers/image_ref.py	`100.00% <100.00%> (ø)`
...ripts/python/tasks/managed/collect_index_images.py	`100.00% <100.00%> (ø)`

... and 2 files with indirect coverage changes

Continue to review full report in Codecov by Harness.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a0c68a7...8896c96. Read the comment docs.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

johnbieren · 2026-06-24T18:52:44Z

/agentic_review

johnbieren · 2026-06-25T13:22:18Z

/agentic_review

qodo-app-for-konflux-ci · 2026-06-25T13:29:28Z

Code Review by Qodo

🐞 Bugs (2) 📘 Rule violations (1) 📜 Skill insights (0)

Context used

✅ Compliance rules (platform): 25 rules

1. run_collect_index_images() hardcodes I/O calls ✗ Dismissed 📘 Rule violation ▣ Testability

Description

The new task logic directly calls filesystem and environment dependencies (file.load_json_dict,
Path.write_text, tekton.require_env) without providing overridable callable keyword-argument
injections. This reduces testability and violates the requirement to inject side-effectful
dependencies with production-safe defaults.

Code

scripts/python/tasks/managed/collect_index_images.py[R145-182]

+def run_collect_index_images(
+    *,
+    data_dir: Path,
+    internal_request_results_file: Path,
+    build_timestamp: str,
+    snapshot_path: Path,
+    index_image_snapshot_result_path: Path,
+) -> None:
+    """Load results, build the snapshot JSON, and write Tekton outputs."""
+    results_path = data_dir / internal_request_results_file
+    if not results_path.is_file():
+        msg = f"internal request results file not found: {results_path}"
+        raise FileNotFoundError(msg)
+
+    logger.info("Loading internal request results from %s", results_path)
+    results = file.load_json_dict(results_path)
+    snapshot = collect_index_image_components(results, build_timestamp)
+
+    snapshot_path.parent.mkdir(parents=True, exist_ok=True)
+    snapshot_path.write_text(json.dumps(snapshot) + "\n", encoding="utf-8")
+    index_image_snapshot_result_path.write_text(SNAPSHOT_FILENAME, encoding="utf-8")
+    logger.info("Wrote index image snapshot to %s", snapshot_path)
+
+
+def main() -> int:
+    """Run the collect-index-images workflow."""
+    data_dir = Path(tekton.require_env("PARAM_DATA_DIR"))
+    run_collect_index_images(
+        data_dir=data_dir,
+        internal_request_results_file=Path(
+            tekton.require_env("PARAM_INTERNAL_REQUEST_RESULTS_FILE"),
+        ),
+        build_timestamp=tekton.require_env("PARAM_BUILD_TIMESTAMP"),
+        snapshot_path=data_dir / SNAPSHOT_FILENAME,
+        index_image_snapshot_result_path=tekton.result_paths_from_env(
+            "RESULT_INDEX_IMAGE_SNAPSHOT_PATH",
+        )[0],
+    )

Evidence
PR Compliance ID 906 requires side-effectful dependencies (I/O, env reads) to be injected as
callable keyword arguments with production-ready defaults. The new managed task script calls these
dependencies directly in run_collect_index_images() and main() without any injectable *_fn
keyword parameters.
Rule 906: Inject dependencies as overridable callable keyword arguments with production-safe defaults
scripts/python/tasks/managed/collect_index_images.py[145-182]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`scripts/python/tasks/managed/collect_index_images.py` performs side-effectful operations (env reads + file I/O) via hardcoded module/method calls. Per compliance, these dependencies must be injected as overridable callable keyword arguments with production-safe defaults.

## Issue Context
This affects `run_collect_index_images()` and `main()`, which currently call `tekton.require_env`, `file.load_json_dict`, and `Path.write_text` directly.

## Fix Focus Areas
- scripts/python/tasks/managed/collect_index_images.py[145-182]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

2. Unstripped index values 🐞 Bug ≡ Correctness

Description

collect_index_image_components() validates target_index/index_image_resolved using .strip() but then
continues processing the unstripped strings, so leading/trailing whitespace can prevent
delivery-repo translation and cause build_tags() to miss bare OCP version tags.

Code

scripts/python/tasks/managed/collect_index_images.py[R101-111]

+        target_index = row.get("target_index")
+        source_index = row.get("index_image_resolved")
+        if not isinstance(target_index, str) or not target_index.strip():
+            msg = f"components[{index}].target_index must be a non-empty string"
+            raise ValueError(msg)
+        if not isinstance(source_index, str) or not source_index.strip():
+            msg = f"components[{index}].index_image_resolved must be a non-empty string"
+            raise ValueError(msg)
+
+        repository, tag = split_target_index(target_index)
+        tags = build_tags(tag, build_timestamp)

Evidence
The task validates stripped values but proceeds with the original strings; both tag matching and
delivery-repo prefix matching are whitespace-sensitive.
scripts/python/tasks/managed/collect_index_images.py[101-125]
scripts/python/tasks/managed/collect_index_images.py[29-34]
scripts/python/helpers/image_ref.py[22-33]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`collect_index_image_components()` checks `target_index.strip()` / `source_index.strip()` for non-emptiness, but then passes the original variables into `split_target_index()`, `build_tags()`, and `translate_delivery_repo()`. If the JSON contains whitespace (common when values come from templating or manual edits), tag parsing and translation can silently fail.

## Issue Context
- `build_tags()` uses a strict `fullmatch()` on the tag string.
- `translate_delivery_repo()` uses `startswith()` against exact prefixes.

## Fix Focus Areas
- scripts/python/tasks/managed/collect_index_images.py[101-123]
- scripts/python/helpers/image_ref.py[22-33]

## Suggested fix
1) In `collect_index_image_components()`, after the type/emptiness checks, normalize:
```py
assert isinstance(target_index, str)
assert isinstance(source_index, str)
target_index = target_index.strip()
source_index = source_index.strip()
```
Then use the normalized variables for `split_target_index()`, logging, and `translate_delivery_repo()`.

2) Optionally harden `translate_delivery_repo()` itself:
- Replace `normalized = repo.replace(...)` with `repo = repo.strip(); normalized = repo.replace(...)` so callers don’t need to remember to trim.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

3. Bad colon tag stripping 🐞 Bug ≡ Correctness

Description

translation_repo_url() removes a tag by splitting on the first ':', which will truncate valid
registry references that include a host port (e.g. registry:5000/repo:tag), yielding an incorrect
rh-registry-repo/registry-access-repo value.

Code

scripts/python/tasks/managed/collect_index_images.py[R37-45]

+def translation_repo_url(translated: list[dict[str, Any]], repo_key: str) -> str:
+    """Return the host/path portion of a translated delivery-repo URL."""
+    for entry in translated:
+        if entry.get("repo") != repo_key:
+            continue
+        url = entry.get("url")
+        if not isinstance(url, str) or not url:
+            return ""
+        return url.split(":", 1)[0]

Evidence
The PR implementation strips tags via a first-colon split, which will truncate host:port references;
existing repo code uses last-colon splitting to avoid this.
scripts/python/tasks/managed/collect_index_images.py[37-46]
scripts/python/helpers/authentication.py[229-240]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`translation_repo_url()` currently returns `url.split(":", 1)[0]`, which drops everything after the *first* colon. For registry references that contain a port (e.g. `registry:5000/ns/repo:tag`), this returns only `registry`, corrupting the computed `rh-registry-repo` / `registry-access-repo`.

## Issue Context
Other code in this repo already uses `rsplit(':', 1)` to remove tag suffixes while preserving host:port.

## Fix Focus Areas
- scripts/python/tasks/managed/collect_index_images.py[37-46]

## Suggested fix
Replace the naive split with parsing that removes a tag only when the colon appears after the last `/`, and strip digests safely:
- Start with `ref = url.split("@", 1)[0]` (drop digest if present)
- Find `last_slash = ref.rfind("/")` and `last_colon = ref.rfind(":")`
- If `last_colon > last_slash`, return `ref[:last_colon]` else return `ref`
This preserves `registry:5000/...` while still stripping `...:tag`.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

4. Non-CheckStepError raises in collect_index_images 📘 Rule violation ≡ Correctness ⭐ New

Description

The managed Tekton task script raises ValueError/FileNotFoundError for expected step-failure
conditions instead of raising tekton.CheckStepError. This can break standardized Tekton failure
handling and result reporting expected by the shared helpers.

Code

scripts/python/tasks/managed/collect_index_images.py[R154-157]

+    results_path = data_dir / internal_request_results_file
+    if not results_path.is_file():
+        msg = f"internal request results file not found: {results_path}"
+        raise FileNotFoundError(msg)

Relevance

⭐ Low
Managed tasks intentionally let generic exceptions propagate; recent merged managed scripts raise
FileNotFoundError/ValueError without CheckStepError.
PR-#841
PR-#813

ⓘ Recommendations generated based on similar findings in past PRs

Evidence
PR Compliance ID 904 requires signaling expected Tekton step failures via
helpers.tekton.CheckStepError (imported here as tekton) rather than ad-hoc exceptions/exits. The
new managed task code raises ValueError and FileNotFoundError directly for validation and
missing-input conditions, which are typical step-failure paths.
Rule 904: Use helpers.tekton.CheckStepError for Tekton step failures
scripts/python/tasks/managed/collect_index_images.py[20-26]
scripts/python/tasks/managed/collect_index_images.py[154-157]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`scripts/python/tasks/managed/collect_index_images.py` signals expected Tekton step failures by raising `ValueError` / `FileNotFoundError` directly. Per the checklist, expected step-failure conditions in Tekton scripts should use `tekton.CheckStepError` so failures are standardized and can be rendered consistently into Tekton results.

## Issue Context
Examples of expected failures here include invalid `target_index` formatting and missing internal-results input files.

## Fix Focus Areas
- scripts/python/tasks/managed/collect_index_images.py[20-26]
- scripts/python/tasks/managed/collect_index_images.py[154-157]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

Previous review results

Review updated until commit bbf98f2

Results up to commit 5a0ab16

🐞 Bugs (2) 📘 Rule violations (0) 📎 Requirement gaps (0) 🎨 UX issues (0) 🔗 Cross-repo conflicts (0) 📜 Skill insights (0)

1. run_collect_index_images() hardcodes I/O calls ✗ Dismissed 📘 Rule violation ▣ Testability

Description

The new task logic directly calls filesystem and environment dependencies (file.load_json_dict,
Path.write_text, tekton.require_env) without providing overridable callable keyword-argument
injections. This reduces testability and violates the requirement to inject side-effectful
dependencies with production-safe defaults.

Code

scripts/python/tasks/managed/collect_index_images.py[R145-182]

+def run_collect_index_images(
+    *,
+    data_dir: Path,
+    internal_request_results_file: Path,
+    build_timestamp: str,
+    snapshot_path: Path,
+    index_image_snapshot_result_path: Path,
+) -> None:
+    """Load results, build the snapshot JSON, and write Tekton outputs."""
+    results_path = data_dir / internal_request_results_file
+    if not results_path.is_file():
+        msg = f"internal request results file not found: {results_path}"
+        raise FileNotFoundError(msg)
+
+    logger.info("Loading internal request results from %s", results_path)
+    results = file.load_json_dict(results_path)
+    snapshot = collect_index_image_components(results, build_timestamp)
+
+    snapshot_path.parent.mkdir(parents=True, exist_ok=True)
+    snapshot_path.write_text(json.dumps(snapshot) + "\n", encoding="utf-8")
+    index_image_snapshot_result_path.write_text(SNAPSHOT_FILENAME, encoding="utf-8")
+    logger.info("Wrote index image snapshot to %s", snapshot_path)
+
+
+def main() -> int:
+    """Run the collect-index-images workflow."""
+    data_dir = Path(tekton.require_env("PARAM_DATA_DIR"))
+    run_collect_index_images(
+        data_dir=data_dir,
+        internal_request_results_file=Path(
+            tekton.require_env("PARAM_INTERNAL_REQUEST_RESULTS_FILE"),
+        ),
+        build_timestamp=tekton.require_env("PARAM_BUILD_TIMESTAMP"),
+        snapshot_path=data_dir / SNAPSHOT_FILENAME,
+        index_image_snapshot_result_path=tekton.result_paths_from_env(
+            "RESULT_INDEX_IMAGE_SNAPSHOT_PATH",
+        )[0],
+    )

Evidence
PR Compliance ID 906 requires side-effectful dependencies (I/O, env reads) to be injected as
callable keyword arguments with production-ready defaults. The new managed task script calls these
dependencies directly in run_collect_index_images() and main() without any injectable *_fn
keyword parameters.
Rule 906: Inject dependencies as overridable callable keyword arguments with production-safe defaults
scripts/python/tasks/managed/collect_index_images.py[145-182]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`scripts/python/tasks/managed/collect_index_images.py` performs side-effectful operations (env reads + file I/O) via hardcoded module/method calls. Per compliance, these dependencies must be injected as overridable callable keyword arguments with production-safe defaults.

## Issue Context
This affects `run_collect_index_images()` and `main()`, which currently call `tekton.require_env`, `file.load_json_dict`, and `Path.write_text` directly.

## Fix Focus Areas
- scripts/python/tasks/managed/collect_index_images.py[145-182]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

2. Bad colon tag stripping 🐞 Bug ≡ Correctness

Description

translation_repo_url() removes a tag by splitting on the first ':', which will truncate valid
registry references that include a host port (e.g. registry:5000/repo:tag), yielding an incorrect
rh-registry-repo/registry-access-repo value.

Code

scripts/python/tasks/managed/collect_index_images.py[R37-45]

+def translation_repo_url(translated: list[dict[str, Any]], repo_key: str) -> str:
+    """Return the host/path portion of a translated delivery-repo URL."""
+    for entry in translated:
+        if entry.get("repo") != repo_key:
+            continue
+        url = entry.get("url")
+        if not isinstance(url, str) or not url:
+            return ""
+        return url.split(":", 1)[0]

Evidence
The PR implementation strips tags via a first-colon split, which will truncate host:port references;
existing repo code uses last-colon splitting to avoid this.
scripts/python/tasks/managed/collect_index_images.py[37-46]
scripts/python/helpers/authentication.py[229-240]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`translation_repo_url()` currently returns `url.split(":", 1)[0]`, which drops everything after the *first* colon. For registry references that contain a port (e.g. `registry:5000/ns/repo:tag`), this returns only `registry`, corrupting the computed `rh-registry-repo` / `registry-access-repo`.

## Issue Context
Other code in this repo already uses `rsplit(':', 1)` to remove tag suffixes while preserving host:port.

## Fix Focus Areas
- scripts/python/tasks/managed/collect_index_images.py[37-46]

## Suggested fix
Replace the naive split with parsing that removes a tag only when the colon appears after the last `/`, and strip digests safely:
- Start with `ref = url.split("@", 1)[0]` (drop digest if present)
- Find `last_slash = ref.rfind("/")` and `last_colon = ref.rfind(":")`
- If `last_colon > last_slash`, return `ref[:last_colon]` else return `ref`
This preserves `registry:5000/...` while still stripping `...:tag`.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

3. Unstripped index values 🐞 Bug ≡ Correctness

Description

collect_index_image_components() validates target_index/index_image_resolved using .strip() but then
continues processing the unstripped strings, so leading/trailing whitespace can prevent
delivery-repo translation and cause build_tags() to miss bare OCP version tags.

Code

scripts/python/tasks/managed/collect_index_images.py[R101-111]

+        target_index = row.get("target_index")
+        source_index = row.get("index_image_resolved")
+        if not isinstance(target_index, str) or not target_index.strip():
+            msg = f"components[{index}].target_index must be a non-empty string"
+            raise ValueError(msg)
+        if not isinstance(source_index, str) or not source_index.strip():
+            msg = f"components[{index}].index_image_resolved must be a non-empty string"
+            raise ValueError(msg)
+
+        repository, tag = split_target_index(target_index)
+        tags = build_tags(tag, build_timestamp)

Evidence
The task validates stripped values but proceeds with the original strings; both tag matching and
delivery-repo prefix matching are whitespace-sensitive.
scripts/python/tasks/managed/collect_index_images.py[101-125]
scripts/python/tasks/managed/collect_index_images.py[29-34]
scripts/python/helpers/image_ref.py[22-33]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`collect_index_image_components()` checks `target_index.strip()` / `source_index.strip()` for non-emptiness, but then passes the original variables into `split_target_index()`, `build_tags()`, and `translate_delivery_repo()`. If the JSON contains whitespace (common when values come from templating or manual edits), tag parsing and translation can silently fail.

## Issue Context
- `build_tags()` uses a strict `fullmatch()` on the tag string.
- `translate_delivery_repo()` uses `startswith()` against exact prefixes.

## Fix Focus Areas
- scripts/python/tasks/managed/collect_index_images.py[101-123]
- scripts/python/helpers/image_ref.py[22-33]

## Suggested fix
1) In `collect_index_image_components()`, after the type/emptiness checks, normalize:
```py
assert isinstance(target_index, str)
assert isinstance(source_index, str)
target_index = target_index.strip()
source_index = source_index.strip()
```
Then use the normalized variables for `split_target_index()`, logging, and `translate_delivery_repo()`.

2) Optionally harden `translate_delivery_repo()` itself:
- Replace `normalized = repo.replace(...)` with `repo = repo.strip(); normalized = repo.replace(...)` so callers don’t need to remember to trim.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

johnbieren · 2026-06-25T13:33:54Z

/agentic_review

johnbieren · 2026-06-25T14:02:11Z

/agentic_review

johnbieren · 2026-06-25T14:14:58Z

/agentic_review

johnbieren · 2026-06-25T14:26:06Z

/agentic_review

seanconroy2021 · 2026-06-25T14:33:57Z

/agentic_review

qodo-app-for-konflux-ci · 2026-06-25T18:49:17Z

PR Summary by Qodo

Add collect_index_images managed task and delivery-repo translation helper
✨ Enhancement 🧪 Tests 🕐 40+ Minutes

Description

• Add a Python collect-index-images managed task to generate Pyxis snapshot JSON.
• Translate Quay delivery-repo image references to public registry URLs.
• Add unit tests covering translation rules, tagging behavior, and snapshot generation.

Diagram

graph TD
  A["Tekton params/env"] --> B["collect_index_images.py"] --> C[("Internal results JSON")]
  B --> D["image_ref.translate_delivery_repo"]
  B --> E[("index_image_snapshot.json")]
  B --> F[("Tekton result file")]

High-Level Assessment

The PR’s approach (porting bash task logic to a typed, unit-tested Python module and reusing a shared translation helper) is the most maintainable option for managed tasks. Keeping the logic in inline bash would reduce reuse and make validation/testing harder; no materially better architectural alternative stands out for this scope.

Files changed (4) +695 / -0

Enhancement (2) +242 / -0

image_ref.pyAdd delivery-repo to public-registry translation helper +55/-0

Add delivery-repo to public-registry translation helper

• Introduces translate_delivery_repo() to map Quay delivery-repo references (including flatpaks and stage/prod) into corresponding registry.redhat.io and registry.access.* URLs. Emits a warning and passes through unknown formats while leaving access URL blank.

scripts/python/helpers/image_ref.py

collect_index_images.pyAdd managed task to build index-image snapshot from internal-request results +187/-0

Add managed task to build index-image snapshot from internal-request results
• Implements a Tekton-invoked script that loads internal-request results, validates schema, derives snapshot tags (including timestamp suffix for bare vX.Y tags), translates delivery repos, and writes index_image_snapshot.json plus a Tekton result containing the snapshot filename.
scripts/python/tasks/managed/collect_index_images.py

Tests (2) +453 / -0

test_image_ref.pyAdd unit tests for delivery-repo translation rules +86/-0

Add unit tests for delivery-repo translation rules
• Adds pytest coverage for translate_delivery_repo() across prod/pending, flatpaks stage/prod, index-image style references, empty input, and unknown-format warning behavior.
scripts/python/helpers/test_image_ref.py

test_collect_index_images.pyAdd unit tests for snapshot generation and Tekton I/O behavior +367/-0

Add unit tests for snapshot generation and Tekton I/O behavior
• Adds comprehensive tests for tag derivation, target_index parsing, delivery-repo field population, imageDigests propagation/defaulting, input validation failures, and run/main behaviors including missing input file handling.
scripts/python/tasks/managed/test_collect_index_images.py

qodo-app-for-konflux-ci · 2026-06-25T19:02:17Z

Code review by qodo was updated up to the latest commit bbf98f2

FilipNikolovski

lgtm once @querti comments are resolved.

This commit adds a python script to replicate the functionality of the inline bash script in the collect-index-images managed task in the catalog repo. The task will be updated to use this python module instead. Assisted-By: Cursor Signed-off-by: Johnny Bieren <jbieren@redhat.com>

johnbieren · 2026-06-29T18:12:26Z

/retest

johnbieren · 2026-06-30T13:40:31Z

/retest

ach912

lgtm

ach912 · 2026-07-01T07:24:35Z

Production Approval Record

Field	Value
Action	APPROVED
Reviewer	@ach912
Timestamp	2026-07-01T07:24:32.895Z

Approved

github-actions Bot added the ai-assisted label Jun 24, 2026

johnbieren force-pushed the release2479 branch from 43a9ba9 to 5a0ab16 Compare June 25, 2026 14:02

qodo-app-for-konflux-ci Bot reviewed Jun 25, 2026

View reviewed changes

Comment thread scripts/python/tasks/managed/collect_index_images.py

johnbieren force-pushed the release2479 branch from 5a0ab16 to bbf98f2 Compare June 25, 2026 17:13

johnbieren marked this pull request as ready for review June 25, 2026 18:44

release-service-bot assigned FilipNikolovski and querti Jun 25, 2026

querti reviewed Jun 26, 2026

View reviewed changes

Comment thread scripts/python/tasks/managed/collect_index_images.py

Comment thread scripts/python/helpers/image_ref.py Outdated

Comment thread scripts/python/helpers/image_ref.py

Comment thread scripts/python/tasks/managed/collect_index_images.py Outdated

johnbieren force-pushed the release2479 branch from bbf98f2 to 234cd90 Compare June 26, 2026 13:13

FilipNikolovski reviewed Jun 29, 2026

View reviewed changes

johnbieren force-pushed the release2479 branch from 234cd90 to 8896c96 Compare June 29, 2026 17:51

querti approved these changes Jun 30, 2026

View reviewed changes

johnbieren added the prod/needs-approval label Jun 30, 2026

davidmogar approved these changes Jun 30, 2026

View reviewed changes

johnbieren requested a review from twaugh June 30, 2026 18:11

ach912 approved these changes Jul 1, 2026

View reviewed changes

ach912 added the prod/approved label Jul 1, 2026

ach912 removed the prod/needs-approval label Jul 1, 2026

johnbieren added this pull request to the merge queue Jul 1, 2026