Conversation
|
@codex review |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
🛡️ Codex Security Review · Automatically triggeredSecurity review completed. No security issues were found in this pull request. Reviewed commit: Only the user who started this review can view the report in Codex. ℹ️ About Codex security reviews in GitHubThis is an experimental Codex feature. Security reviews are triggered when:
Once complete, Codex will leave suggestions, or a comment if no findings are found. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1de182b5a4
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| runs-on: ubuntu-latest | ||
| timeout-minutes: 5 | ||
| steps: | ||
| - uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0 |
There was a problem hiding this comment.
Because this workflow sets top-level permissions: {}, the new Quality Gate inherits a GITHUB_TOKEN without contents: read; unlike the existing test and mutation jobs, it does not override that permission before invoking actions/checkout. The gate can therefore fail at checkout on every trigger before validating the shard reports, so add job-level contents: read or avoid checking out the repository.
Useful? React with 👍 / 👎.
| @@ -0,0 +1,118 @@ | |||
| """Require every runtime module in exactly one successful mutation shard.""" | |||
There was a problem hiding this comment.
This adds a custom enforcement mechanism without documenting the native-tool gap it fills, while maintainers/mutation-testing.md still says poe quality mutates changed modules and that CI runs the mutation task. Update the maintainer guidance to describe the new full/sharded orchestration and why Mutmut/GitHub's native configuration cannot enforce cross-shard completeness; the repository explicitly requires that justification for custom checks.
AGENTS.md reference: AGENTS.md:L3-L5
Useful? React with 👍 / 👎.
|
@codex review |
|
Codex Review: Didn't find any major issues. Another round soon, please! Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
🛡️ Codex Security Review · Automatically triggeredSecurity review completed. No security issues were found in this pull request. Reviewed commit: Only the user who started this review can view the report in Codex. ℹ️ About Codex security reviews in GitHubThis is an experimental Codex feature. Security reviews are triggered when:
Once complete, Codex will leave suggestions, or a comment if no findings are found. |
583e5f2 to
6c94b2f
Compare
6c94b2f to
22c27c8
Compare
22c27c8 to
c40e717
Compare
|
@codex review |
🛡️ Codex Security Review · Automatically triggeredSecurity review completed. No security issues were found in this pull request. Reviewed commit: Only the user who started this review can view the report in Codex. ℹ️ About Codex security reviews in GitHubThis is an experimental Codex feature. Security reviews are triggered when:
Once complete, Codex will leave suggestions, or a comment if no findings are found. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: ed1a6d0dc2
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| requires both through `Quality Gate`. The weekly `poe mutation-full` task audits | ||
| every handwritten runtime module without a debt baseline. | ||
| `uv run --locked poe quality` runs all native checks and mutates every handwritten | ||
| runtime module. CI runs `checks` and divides `mutation-full` across eight jobs; |
There was a problem hiding this comment.
Update this to say 23 jobs: the CI matrix defines shards 0–22 and the Quality Gate explicitly requires 23 reports. Describing eight jobs makes the new maintainer guidance internally contradictory and gives maintainers the wrong topology when investigating capacity, timeouts, or future shard-count changes.
Useful? React with 👍 / 👎.
|
@codex review |
🛡️ Codex Security Review · Automatically triggeredSecurity review completed. No security issues were found in this pull request. Reviewed commit: Only the user who started this review can view the report in Codex. ℹ️ About Codex security reviews in GitHubThis is an experimental Codex feature. Security reviews are triggered when:
Once complete, Codex will leave suggestions, or a comment if no findings are found. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: ed1a6d0dc2
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| strategy: | ||
| fail-fast: false | ||
| matrix: | ||
| shard: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22] |
There was a problem hiding this comment.
With the current 24-module inventory, 23 shards work only because shard 0 receives both export-only __init__.py and one mutatable module. If a PR removes any handwritten runtime module, shard 0 receives only __init__.py; scripts.mutation_results then reports “No mutants were tested,” so the mandatory mutation job fails even though every remaining mutatable module was covered. Derive the shard topology from the inventory or otherwise ensure every shard includes a mutatable module.
Useful? React with 👍 / 👎.
| async with asyncio.timeout_at(deadline): | ||
| return await self._queue.get() |
There was a problem hiding this comment.
When batch_window_seconds is zero, or the deadline has otherwise elapsed, asyncio.timeout_at() does not prevent an already-ready Queue.get() from succeeding because timeout cancellation is only scheduled for the next loop iteration. If jobs are queued while an earlier fetch is blocked, the worker therefore batches those queued jobs together despite a zero batch window—for example, three jobs can produce fetches (1,) and (2, 3) instead of three separate fetches. Restore an explicit expired-deadline check before dequeuing.
Useful? React with 👍 / 👎.
1 participant
Python's PR mutation job previously covered changed modules and locks, leaving untouched handwritten runtime code outside the required gate. The scheduled full audit also could not start because Poe treated
MUTATION_FULL=1as an executable. This PR makespoe qualityrun the complete mutation set through Poe's native shell task.CI distributes all 24 tracked handwritten runtime modules across 23 Mutmut jobs. Each job uploads native outcome metadata; the required Quality Gate verifies that every report exists, every shard succeeds, and each module appears exactly once. Surviving, uncovered, crashed, timed-out, incomplete, or empty results fail. The source inventory comes from Git at the tested commit, so new modules enter automatically. Pinned Pyrefly checks every handwritten module before running each mutant; test-killed and statically invalid mutants remain distinct outcomes.
The URL and realtime fetch worker cleanup removes equivalent mutations and covers cancellation, backpressure, callback cleanup, and query preservation with deterministic tests. The URL rewrite was compared with the previous implementation over 100,000 generated inputs. A bounded callback test avoids counting timeout machinery as a worker future on Python 3.11.
Local validation on current
main:poe checks(2,262 tests; 100% of 5,193 runtime lines and 1,316 branches), Ruff, mypy, basedpyright, package builds, isolated optional-dependency installs, installed-package typing (845/845 exported symbols), shared contract dry-run, and policy inventory. The local scoped mutation run killed 307 mutants and statically rejected 444, with no failures. Exact-head CI covered all 24 handwritten runtime modules across 23 shards: 3,247 mutants killed, 5,136 statically invalid, no survivors, uncovered mutants, timeouts, or crashes.__init__.pyhas no mutable operators.