agentleFS
vortexgpgpuvortexgpgpu/vortex/.agents/skills/code-review/SKILL.md

code-review

Skill2.4k starsChanged 7 days agoApache-2.0View on GitHub

Review a Vortex diff (a commit, a branch range or a pull request) against the repository's own rules, classify its change impact (R0-R3) and report findings by severity (F0-F3). Use before landing any change, when asked to review a PR or a commit range, and as the independent reviewer an R3 change requires. Static review; it never edits, commits or pushes.

194 other projects use a version of this.

Trending

More talk lately.thin dataabout the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 1 mentionApr 2026: 0 mentionsMay 2026: 1 mentionJun 2026: 0 mentionsJul 2026: 1 mentionAug 2026: 0 mentionsSep 2026: 1 mentionOct 2026: 1 mentionNov 2026: 1 mention
Dec 2025Jun 2026Nov 2026
Mentions a month of the vortexgpgpu/vortex repository that our weekly crawl found: only 13 in all, so this is a hint, not a trend. There is no adoption data for this file yet.

Most recently taken up by

Who's talking about vortexgpgpu/vortex

The biggest 5 of 13 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 13

Talked about on GitHub and Hacker News (13 times).

On GitHub · 11

6 more on GitHub

On Hacker News · 2

Used with

Projects with it also use close-spec, design-with-images and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • dzhngSkill

    close-spec

    When a spec under specs/<feature>/ is done shipping (a write-spec build, or any planned task that produced a spec), archive it to specs/done/ and rewrite it from a build-plan into a durable rationale — the why, the principles, the invariants — pointing back to the real code for the how. Use when implementation has landed and the plan no longer matches what shipped, or the user says a feature/spec is finished. Pairs with [write-spec](../write-spec/SKILL.md) (the plan this closes) and [review](../review/SKILL.md) (run before closing).

    Also in 5 projects that use this

  • dzhngSkill

    design-with-images

    Explore UI and visual design with image generation, then iterate the real implementation against the user-selected concept. Use when brainstorming visual directions, offering mockup variants, or implementing an approved generated design; do not stop at the mockups or the first code approximation.

    Also in 5 projects that use this

  • PydanticSkill

    add-new-model

    Add support for a newly-released language or image generation model in pydantic-ai (e.g. openai:gpt-5.6, anthropic:claude-sonnet-5, openai:gpt-image-2). Use when a provider ships a new model id and you need to wire literals, profile flags, adapters, and tests to recognize it. Handles SDK-lag, gateway list conventions, capability probing, and direct image-model geometry.

    Also in 9 projects that use this

  • transitive-bullshitSkill

    compare-screenshots

    Compare screenshots against the intended design, distinguishing approved references from historical baselines. Use for iterative implementation matching, before/after visual review, objective image telemetry, or checking a lone capture for flat, empty, or badly framed content.

    Also in 5 projects that use this

  • Microsoft AzureSkill

    azsdk-common-apiview-feedback-resolution

    Analyze and resolve APIView review feedback on Azure SDK PRs. **UTILITY SKILL**. USE FOR: APIView comments, API review feedback, SDK API surface changes. DO NOT USE FOR: general code review, non-APIView feedback. INVOKES: azure-sdk-mcp:azsdk_apiview_get_comments, azure-sdk-mcp:azsdk_customized_code_update, azure-sdk-mcp:azsdk_typespec_delegate_apiview_feedback, azure-sdk-mcp:azsdk_run_typespec_validation, azure-sdk-mcp:azsdk_package_generate_code.

    Also in 6 projects that use this

  • Microsoft AzureSkill

    azsdk-common-generate-sdk-locally

    Generate and finalize Azure SDKs locally from TypeSpec, including build, checks, tests, changelog, metadata, and version updates by default. WHEN: "generate SDK locally", "build SDK", "run SDK tests", "run CI checks", "validate package", "run checks", "update changelog", "fix SDK build errors", "resolve SDK generation errors", "customize TypeSpec", "rename SDK client", "rename SDK model", "hide operation from SDK", "fix analyzer errors", "resolve customization drift", "create subclient", "update metadata", "update version". INVOKES: azsdk_verify_setup, azsdk_package_generate_code, azsdk_package_build_code, azsdk_package_run_check, azsdk_package_run_tests, azsdk_customized_code_update, azsdk_package_update_changelog_content, azsdk_package_update_metadata, azsdk_package_update_version.

    Also in 6 projects that use this

  • +4 more from dzhng
  • +2 more from Azure

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 184 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read these in full before reading the diff.

    line 13
  2. Distinctive:

    Do not report style preferences, naming taste or anything the documents do not require.

    line 72
  3. Distinctive:

    Verify every finding against the code at the head of the range before reporting it.

    line 88
  4. Distinctive:

    Run heavy builds (rtlsim, synthesis) to settle a review question without being asked.

    line 99

“Common” is measured: kept by at least 6 in 10 of the 184 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Reviewing a Vortex change

The policy is docs/review/code-review-policy.md. This skill is how to apply it.

1. Load the rules

Read these in full before reading the diff. They are not auto-loaded.

  • AGENTS.md
  • docs/coding_guidelines_verilog.md, docs/coding_guidelines_cpp.md
  • docs/bug_fixes.md
  • docs/review/code-review-policy.md

For RTL hunks also follow the rtl-review skill; for sim/simx/ hunks, the simx-review skill.

2. Classify

python3 ci/change_impact.py classify --base <base> --head <head>
python3 ci/change_impact.py explain  --base <base> --head <head>   # per file, with rules

The class decides the depth: R0 needs no review, R1 a read of the change, R2 and R3 every changed hunk. Each contract finding and each evidence item the classifier prints is something the review must confirm or refute. The classifier is advisory: if the diff carries a contract change it missed, say so and treat the change as R3.

3. Review the diff

Review the diff, not the files around it. Comment on unchanged code only where the diff makes it wrong. For each hunk ask:

  • Does it do what the commit message says, on every configuration the toml and ci/regression.sh allow, at XLEN 32 and 64?
  • What else consumes what it changed (an interface, a struct, a CSR, a DCR, a header, a testcase, a baseline)? Did those consumers change with it?
  • Can it move cycles? Then where is the SimX side, and which model_parity case shows it?
  • Does it fix a root cause, or move a symptom (docs/bug_fixes.md §2)?

For a range or a branch, also review the final state of every file that several commits touched: the end state matters more than each step.

A large or cross-subsystem diff goes to a fresh reviewer that did not write it, with the diff and the rule documents; an R3 change always does.

4. Hard rules: always F2 or F3

rulewhy
golden data (ci/baselines/**, reference images) changed without the update command in the commit message, or re-recorded by an agent or CIa hand edit absorbs a regression; only a human re-records golden data
a test silenced: known_issue:, wider tolerance:/threshold, smaller args:, a removed case, to make a failure go awayfix the bug instead
an existing test case changed or removed, a golden value moved toward a looser bar (more cycles, lower Fmax, more area), or a gate threshold widened, without the stated, measured reason and the maintainer's approvalthese change what CI proves (R3: test-coverage, golden-values, gate-thresholds)
a bug papered over (fallback, clamp, delay, retry, disabled feature) without a PATCH:/WORKAROUND: label and a follow-updocs/bug_fixes.md §3
a change that claims to fix a perf, area or timing regression but does not touch the commit that caused it (its code, parameter or default, or a revert); or a baseline re-record without the case → commit → mechanism → class table; or a justification with no proof that the cost is inherentdocs/bug_fixes.md §6: a compensating gain elsewhere hides the regression
sw/kernel or sw/runtime references hw/ or sim/, or the reverselayer isolation; shared ABI goes in sw/common
a Verilog interface added to RTL_PKGSRTL_PKGS is VX_*_pkg.sv only
a comment names a task, PR, issue, phase or "the fix", or names the other layer's internalscomments explain why, and rot otherwise
RTL timing changed without SimX, or the reverse, with no parity evidenceSimX is the RTL's timing model
a toolchain pin moved before the prebuilt tag existsCI downloads binaries at the pinned tag
a switch that is parsed and ignoredworse than no switch
a workflow exposes a self-hosted runner or secrets to a forktrust boundary

Do not report style preferences, naming taste or anything the documents do not require.

5. Report

One entry per finding:

F2 high static-proof  <sha> path/to/file.sv:123
What: <the defect, one sentence>
Scenario: <the configuration and input that make it fail, and what is observed>
Fix: <the smallest correct change>
  • Severity F0-F3 and confidence as defined in the policy. Evidence type is static-proof, reasoned-risk or missing-test.
  • Verify every finding against the code at the head of the range before reporting it. If it no longer holds there, drop it or say it was fixed later.
  • State what you did not review and what evidence you could not see. A test you did not see run is not run, never a pass.
  • End with the class, the obligations still open, and a verdict: land, land after the listed fixes, or do not land.

6. Never

  • Edit, commit, push, or run checkout/reset/stash/clean while reviewing: the trees are shared.
  • Run heavy builds (rtlsim, synthesis) to settle a review question without being asked; say what would settle it instead.
  • Claim a result you did not see.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.