review-cudf
rapidsai/cudf/.agents/skills/review-cudf/SKILL.md
Use this skill to review GitHub pull requests for cudf
What's in it
- Review cuDF Pull Request
- Review Checklist
- Reference Material
- Output Format
--- name: review-cudf description: Use this skill to review GitHub pull requests for cudf --- Use this skill when the user invokes `/review-cudf` with: - a cudf GitHub PR link - currently checked out cudf PR - specified cudf code changes or a diff cudf GitHub repository is located at: https://github.com/NVIDIA/cudf # Review cuDF Pull Request 1. **Fetch PR metadata and diff** ```bash gh pr view <PR_NUMBER> --repo NVIDIA/cudf --json title,body,files,additions,deletions,baseRefName,headRefName gh pr diff <PR_NUMBER> --repo NVIDIA/cudf ``` Hint: Check if `GH_TOKEN` (or GitHub CLI auth) is already configured in the environment (for example via your secret manager) so `gh` can authenticate and bypass rate limits; do not run `gh auth token` from within the agent. If `gh` auth is unavailable, fall back to GitHub's raw diff/patch URLs, `git fetch` of the PR ref, unauthenticated GitHub REST API with `curl`, or any other available methods. 2. **Fetch review comments already posted** for context on what's already been suggested and need not be repeated. 3. **Read the Developer Guide** (`cpp/doxygen/developer_guide/DEVELOPER_GUIDE.md`) — it is the authoritative reference for libcudf conventions. All rules in the guide apply during review. The checklist below calls out the most review-relevant rules and adds items **not** covered by the guide. 4. **Analyze the changes** against the checklist below, reading relevant source files as needed for context. 5. **Produce a structured review** using the output format at the bottom. 6. **Dump the structured review** to `.agents/reviews/<PR_NUMBER>/review.md` --- ## Review Checklist For the detailed review checklist, read these files: - **C++/CUDA**: `cpp/REVIEW_GUIDELINES.md` - **Python**: `python/REVIEW_GUIDELINES.md` - **Developer Guide**: `cpp/doxygen/developer_guide/DEVELOPER_GUIDE.md` - **Testing Guide**: `cpp/doxygen/developer_guide/TESTING.md` - **Benchmarking Guide**: `cpp/doxygen/developer_guide/BENCHMARKING.md` - **Documentation Guide**: `cpp/doxygen/developer_guide/DOCUMENTATION.md` Use all applicable checklist items from those files when reviewing the PR. The Developer Guide is the authoritative reference for libcudf conventions — all rules in the guide apply during review. --- ## Reference Material | Topic | Path | |-------|------| | Developer guide | `cpp/doxygen/developer_guide/DEVELOPER_GUIDE.md` | | Testing guide | `cpp/doxygen/developer_guide/TESTING.md` | | Benchmarking guide | `cpp/doxygen/developer_guide/BENCHMARKING.md` | | Documentation guide | `cpp/doxygen/developer_guide/DOCUMENTATION.md` | | Profiling guide | `cpp/doxygen/developer_guide/PROFILING.md` | Online libcudf API docs if needed: https://docs.nvidia.com/cudf/latest/libcudf/ --- ## Output Format Structure your review as follows: ```markdown ## PR Review: <PR title> **PR:** <link> **Summary:** <1-2 sentence summary of what the PR does> ### Findings #### Critical - **[file:line]** Description of issue that must be fixed before merge. #### Suggestions - **[file:line]** Description of improvement to consider. #### Nits - **[file:line]** Minor style or formatting issue. Keep these minimal, don't suggest adding comments around every line of code or obvious logic. #### Highlights - Highlight well-written code, good test coverage, or clever solutions. ### Verdict One of: **Approve**, **Request Changes**, or **Comment** With a brief justification. ``` If there are no findings in a category, omit that category.
More agent context in rapidsai/cudf
9 other files this repository gives its agents.
Skill
- build-test-cudf-java.agents/skills/build-test-cudf-java/SKILL.md
- build-test-cudf.agents/skills/build-test-cudf/SKILL.md
- debug-cudf-pandas.agents/skills/debug-cudf-pandas/SKILL.md
- perf-compare-cudf.agents/skills/perf-compare-cudf/SKILL.md
- reproduce-ci.agents/skills/reproduce-ci/SKILL.md
- review-cudf-polars-expressions.agents/skills/review-cudf-polars-expressions/SKILL.md
- update-narwhals-testing.agents/skills/update-narwhals-testing/SKILL.md
- accelerated-computing-cudfskills/accelerated-computing-cudf/SKILL.md
- cudf-query-engine-builderskills/cudf-query-engine-builder/SKILL.md
Also found in one other repository
The same file, byte for byte, in the weekly crawl of public GitHub.
- NVIDIA/cudf9.8k
Discussion
Evidence it works
Elsewhere, people who say they tested it. Read them for what they measured.
- Dec 8, 2021 · GitHub · NVIDIA/cudf · move benchmark input generation to GPU for CONTIGUOUS_SPLIT_BENCH
Talked about elsewhere, linking to rapidsai/cudf
- cuDF – GPU DataFrame Library
Hacker News · 107 points · 29 comments · Jun 2, 2024
- GPU processing - cuDF install problem (O/S or hardware issue?)
Stack Overflow · 8 votes · Aug 27, 2020
- cuDF – GPU DataFrame Library (@github.com) Main Link | Discussion
Bluesky · @newsycombinator.bsky.social · 1 likes and reposts · Jun 3, 2024
- Support the __cuda_array_interface__ protocol
GitHub · tensorflow/tensorflow · 50 reactions · 29 comments · May 26, 2019
- CuDF – GPU DataFrame Library
Hacker News · 1 points · Jun 1, 2024
- ERROR: Could not find a version that satisfies the requirement dask-cudf (from versions: none)
Stack Overflow · 4 votes · 2 answers · Oct 28, 2020
- Add `streaming_groupby` for stateful streaming aggregation
GitHub · NVIDIA/cudf · 9 reactions · 16 comments · Mar 25, 2026
- RAPIDS pip installation issue
Stack Overflow · 3 votes · Jun 29, 2023
Did it work?
Say what you used it for and what you changed. People and their agents can both post here.
No reports yet. Be the first to say whether it worked.
Your agents can post too, on your behalf: the MCP tool registry_write, action report. How to connect one.

