Bump @openai/codex-sdk from 0.146.0 to 0.158.0 - #21
Closed
dependabot[bot] wants to merge 49 commits into
Closed
dependabot[bot] wants to merge 49 commits into
dependabot[bot] wants to merge 49 commits into
Conversation
The repo documented the CLI but never said what CellForge is for or what it achieved. This makes the paper's contribution legible from the front page and adds the project infrastructure a research tool needs to take contributions. README - Lead with the result: PCC 0.9883 on Adamson, MSE_DE 0.1736 on Norman, first in 14/15 blinded judge comparisons, 6 datasets across 3 modalities. - Map the three paper stages (Formulation/Ideation/Execution) onto the package layout, and document the coordination score and its convergence criterion. - Summarise the six generated architectures, the benchmark datasets with GEO accessions, per-run token and GPU cost, and the measured failure-mode distribution. - State the weak results (0.535 DEG recall, 0.420 protein recall) rather than omit them. docs/ - QUICKSTART, ARCHITECTURE, RESULTS, MODELS, DATASETS, FAQ, ROADMAP. Infrastructure - CI: pytest on Python 3.9-3.12, lint, sdist/wheel build, CITATION.cff validation, secret scan. requirements-ci.txt carries the minimal set the 43 tests actually need, so CI skips the full scientific stack. - Issue templates (bug, bad research plan, feature), PR template, dependabot. - CONTRIBUTING, CODE_OF_CONDUCT, SECURITY, CHANGELOG. - pyproject.toml for the build backend and black/isort/ruff/mypy/pytest config. Package metadata stays in setup.py to avoid a second source of truth for dependencies. Data and examples - scripts/download_datasets.py fetches all six benchmark datasets from the scPerturb Zenodo records with md5 verification. Filenames, sizes, and checksums were read from the live Zenodo API; one download was verified end to end. - examples/ carries a task description per benchmark. CITATION.cff listed "CellForge authors" and no paper. It now carries the full author roster and arXiv:2508.02276 as the preferred citation. Verified: 43/43 tests pass, twine check passes, CITATION.cff validates against schema 1.2.0, all .github YAML parses, no broken relative links across the 15 markdown files.
Bumps [actions/checkout](https://github.com/actions/checkout) from 4 to 7. - [Release notes](https://github.com/actions/checkout/releases) - [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md) - [Commits](actions/checkout@v4...v7) --- updated-dependencies: - dependency-name: actions/checkout dependency-version: '7' dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com>
Bumps [@openai/codex-sdk](https://github.com/openai/codex/tree/HEAD/sdk/typescript) from 0.116.0 to 0.146.0. - [Commits](https://github.com/openai/codex/commits/rust-v0.146.0/sdk/typescript) --- updated-dependencies: - dependency-name: "@openai/codex-sdk" dependency-version: 0.146.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com>
Updates the requirements on [referencing](https://github.com/python-jsonschema/referencing) to permit the latest version. - [Release notes](https://github.com/python-jsonschema/referencing/releases) - [Changelog](https://github.com/python-jsonschema/referencing/blob/main/docs/changes.rst) - [Commits](python-jsonschema/referencing@v0.30.2...v0.37.0) --- updated-dependencies: - dependency-name: referencing dependency-version: 0.37.0 dependency-type: direct:production ... Signed-off-by: dependabot[bot] <support@github.com>
A run costs roughly 80k prompt / 400k completion tokens plus 4-8 GPU-hours. Nothing in the repository showed what you get for that, so the decision to try CellForge had to be made blind. examples/outputs/adamson_crispri/ is a complete worked bundle laid out the way a run leaves one: task analysis report, research plan (md/json/mmd), and the generated training script with its metrics and verification output. On provenance, stated per file in PROVENANCE.md and in every README that points at the bundle. This was assembled without an LLM provider key, without the Codex CLI and without the dataset, so it is not the transcript of a live run: - task_analysis_report.md and research_plan.* are hand-written to match the schemas the code actually serialises, section for section (TaskAnalysisReport.to_markdown, refinement.py's five top-level plan keys) - result.py is real, working code. It passes the repository's own deterministic verifier: file, compile, --help, --selftest, and metrics.json structure - metrics.json and verification.json are the genuine output of really executing result.py and really running CodeGenerationVerifier, on synthetic data The synthetic numbers (pcc 0.618, pcc_de 0.218) are a smoke test and are not comparable to the paper's 0.9883 on real Adamson data. PROVENANCE.md says so, explains why genome-wide PCC is inflated on real data, why r2_de is legitimately negative, and notes that the synthetic generator is sympathetic to the architecture it demonstrates. result.py is not a sketch. It conditions the perturbation on the target gene's embedding tied to the decoder output weights, rather than a free lookup table, because held-out rows in a lookup table never receive gradient and the unseen-perturbation split is then unsolvable in principle. torch/numpy/anndata are imported lazily so --help and --selftest run in a bare interpreter. scripts/export_example_run.py packages a genuine run into this layout, scrubbing API keys, absolute paths, usernames and hostnames, and aborting rather than writing a file that still matches a credential pattern. A real bundle should replace this one; CONTRIBUTING.md now asks for exactly that. .gitignore needed an explicit !examples/outputs/ because `outputs/` matched it and git will not descend into an excluded directory. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
…ons/checkout-7 Bump actions/checkout from 4 to 7
…i/codex-sdk-0.146.0 Bump @openai/codex-sdk from 0.116.0 to 0.146.0
…e-0.37.0 Update referencing requirement from >=0.30.2 to >=0.37.0
--- updated-dependencies: - dependency-name: "@openai/codex-sdk" dependency-version: 0.158.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com>
Contributor
Author
LabelsThe following labels could not be found: Please fix the above issues or remove invalid values from |
Contributor
Author
|
OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting If you change your mind, just re-open this PR and I'll resolve any conflicts on it. |
dependabot
Bot
deleted the
dependabot/npm_and_yarn/openai/codex-sdk-0.158.0
branch
October 2, 2026 16:37
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Bumps @openai/codex-sdk from 0.146.0 to 0.158.0.
Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting
@dependabot rebase.Dependabot commands and options
You can trigger Dependabot actions by commenting on this PR:
@dependabot rebasewill rebase this PR@dependabot recreatewill recreate this PR, overwriting any edits that have been made to it@dependabot show <dependency name> ignore conditionswill show all of the ignore conditions of the specified dependency@dependabot ignore this major versionwill close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this minor versionwill close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this dependencywill close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)