Skip to content
Skills Library
Programming · Review & audit

code-review

/code-review

Use when reviewing a pull request or a diff in the neilotoole/sq repository. Carries the repo conventions a reviewer cannot infer from the diff itself: prose and spelling rules, the split between what lint catches and what it does not, test gating, generated fixtures that must not be hand edited, and commit and PR requirements.

Repository
sq ↗
Licence
MIT
Last repository change
25.10.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
2
Deprecation

No signals detected. Repository changed on 25.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when reviewing a pull request or a diff in the neilotoole/sq repository. Carries the repo conventions a reviewer cannot infer from the diff itself: prose and spelling rules, the split between what lint catches and what it does not, test gating, generated fixtures that must not be hand edited, and commit and PR requirements.

Original author description

Use when reviewing a pull request or a diff in the neilotoole/sq repository. Carries the repo conventions a reviewer cannot infer from the diff itself: prose and spelling rules, the split between what lint catches and what it does not, test gating, generated fixtures that must not be hand edited, and commit and PR requirements.

How it works

  1. Highest-value checks
  2. What lint does and does not catch
  3. Flaky tests
  4. Generated fixtures
  5. Test evidence
  6. Other conventions
Original SKILL.md headings · 85 lines, 652 words
SKILL.md preview
---
name: code-review
description: >-
  Use when reviewing a pull request or a diff in the neilotoole/sq repository.
  Carries the repo conventions a reviewer cannot infer from the diff itself:
  prose and spelling rules, the split between what lint catches and what it
  does not, test gating, generated fixtures that must not be hand edited, and
  commit and PR requirements.
license: MIT
metadata:
  homepage: https://sq.io
  version: "0.1.0"
---

# code-review

Review conventions for [`sq`](https://github.com/neilotoole/sq). The canonical
source is [`AGENTS.md`](../../../AGENTS.md); this is the reviewer-facing subset.

Spend review effort on what a general-purpose reviewer would miss. Ordinary Go
correctness, nil handling, and error wrapping are already covered without this
skill. What follows is repo knowledge that is invisible in a diff.

## Highest-value checks

| Check                       | Flag when                                                                                                                                             |
| --------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Em dashes**               | A `—` or `–` appears in prose, godoc, a code comment, markdown, or a YAML comment. Use a period, comma, parentheses, or ": ". Ranges use `-` or "to". |
| **US English**              | British spelling in prose or comments: "honours", "colour", "behaviour", "optimise".                                                                  |
| **Envar**                   | "env var", "env-var" or "environment variable" in prose. The repo term is "envar".                                                                    |
| **`godot`**                 | A comment block whose last line does not end with a period.                                                                                           |
| **`require` over `assert`** | New test code uses `assert.*` without needing to report several independent failures in one run.                                                      |
| **Skipped flaky test**      | A `t.Skip` is added because a test sometimes fails. See [Flaky tests](#flaky-tests).                                                                  |
| **AI attribution**          | A commit message or PR description contains "Generated with", a co-author trailer, or any Claude / AI attribution.                                    |
| **Generated fixtures**      | `site/static/testdata/` is edited by hand. See [Generated fixtures](#generated-fixtures).                                                             |

Those first three do not apply to code itself: string literals, test fixtures
and sample data are exempt. An em dash inside a `testdata` CSV is data, not
prose.

## What lint does and does not catch

`make lint` runs golangci-lint, shellcheck, `dprint check` and biome. It does
**not** catch everything, so these need a human or an agent reviewer:

- **Import grouping.** Enforced by a separate CI step
  (`scripts/fmt-go-imports.sh`). `go build`, `go vet` and golangci-lint all
  pass on wrongly grouped imports. `make fmt` fixes it, and it must be run
  before `make lint`.
- **Prose style.** No linter checks em dashes, US spelling, or "envar".
- **Workflow formatting.** `actionlint` validates workflow syntax, not `dprint`
  style. A workflow can pass `actionlint` and still fail the `Format` CI job.
  Any touched `.yml`, `.json` or `.toml` needs `make fmt`.

## Flaky tests

Do not skip a flaky test; find the root cause and fix it. A `t.Skip` on an
intermittent failure removes the signal and leaves the cause to resurface
against whatever test loses the race next.

Gating on a real precondition is not skipping. These are correct:

- `tu.SkipShort(t, true)` for a test that needs a live database.
- `tu.SkipNoNetwork(t)` for a test that deliberately uses a real remote host.
- The envar checks behind the driver test handles.

Each states what the test requires. A skip added because a test sometimes
failed states nothing. If the cause cannot be fixed in the same change, the PR
should open an issue with the failure output rather than silence the test.

## Generated fixtures

`site/static/testdata/` is generated from the canonical in-repo fixtures by
`go run ./test/fixtures/internal/gentestdata`, and `test/fixtures` guards it
against drift. A PR that edits those files by hand, or that changes a canonical
fixture without regenerating, should be flagged.

Editing a Sakila fixture also changes documented query output, so check whether
`site/content` still matches.

## Test evidence

- A pipe masks a command's exit code. `go test ./... | tail` reports success
  even when the suite fails. Capture the status before piping.
- Server-backed driver tests skip silently when the engine's `SQ_TEST_SRC__*`
  envar is unset, so "tests pass" from a machine without the `sakiladb`
  containers proves less than it appears to.

## Other conventions

- **CHANGELOG.** Work in progress goes under `## Unreleased`. Changes confined
  to `site/` need no entry.
- **Markdown.** Wrap at 100 characters where feasible; `dprint` formats it.
- **Commit messages.** Imperative mood, subject under roughly 70 characters,
  body for the why.
- **Branch names.** `feature/`, `fix/` or `chore/`, plus `gh<ISSUE>-` when a
  GitHub issue is linked.

Versions

RepositoryScoreLicence
This page · neilotoole/sq90MIT
microsoft/aspire100MIT
kdlbs/kandev100AGPL-3.0
dotnet/runtime100MIT
drosam/agents100MIT
ggml-org/llama.cpp100MIT
ShamelesAbyss/Nyx100Apache-2.0
dzhng/jevgrep100MIT
iuliandita/skills100MIT

Database snapshot: 13.11.2026 · package version 2531dcc04677 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90review
    Review a pull request or diff for this repository against otari's project standards. Use when asked to review a PR, inspect a diff, or prepare review feedback.
  • 100frontend-reviewer
    Use when reviewing a frontend pull request or diff, auditing a frontend repo for gaps, or checking whether code has drifted from the project's documented conventions.
  • 100check-diff
    Check the current changes against this repo's own agent rules (CLAUDE.md, AGENTS.md, Cursor rules, Copilot instructions) before committing. Use when the user asks to check, review or verify their changes or diff against the repo's rules or conventions, or before a commit or pull request.
  • 100pg-code-review
    Use when reviewing a local development change or pull request — compare the original requirements and actual scope with owning contracts, standing conventions, defensive patterns, documentation and relevant evidence.
  • 100review-pr
    Review pull requests for system1-agents with high-confidence, evidence-based feedback. Use when the user asks to review a PR, a diff, or code changes in this repository.
  • 90repo-review
    Use when reviewing a pull request or a change in this repository — orients the reviewer to this repository's standards (AGENTS.md conventions, decision log, gates) and the review-specific checks that code alone cannot show

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.