Repository navigation
Rename Narrative Contracts to MateProbe by Mate4B - #2
Conversation
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. |
![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: 6b35931034
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".
|
|
||
|
|
||
| def stage(source: Path, output: Path) -> None: | ||
| manifest = json.loads(Path(__file__).with_name("release-a4.json").read_text()) |
There was a problem hiding this comment.
When the PyPI verification job runs, this unconditional read raises FileNotFoundError because scripts/release-a4.json is not present anywhere in the commit (only the a3 manifest is tracked). Consequently every a4 dry run or publication attempt stops before artifact verification; commit the four-artifact hash manifest referenced here.
AGENTS.md reference: AGENTS.md:L26-L29
Useful? React with 👍 / 👎.
Rename the project to MateProbe by Mate4B, retaining MIT licensing and frozen research artifacts. Distributions are now
mateprobeandpytest-mateprobe; imports, CLI, pytest fixture and report flag follow the new name. Includes a migration guide and separates current installed-package checks from historical a2/a3 trials.Built-in wire identities remain stable so configuration digests and replay results do not change. The frozen real-output generator runs unchanged in an isolated compatibility process; CI compares its reports permitting only the declared runtime version transition.
Validation: 155 tests; Ruff and strict mypy; 8/8 selected source mutants; original and expanded synthetic campaigns; both captured-output replays and real-output mutation comparison; built wheel installation, consumer docs, CLI, pytest and launch regression checks. No frozen benchmark data or previous release files changed.