Move Code Quality Checker
Reviews Move packages against the official Move Book Code Quality Checklist and returns actionable, prioritized feedback.
Pure prompt-based skill: no scripts, no network calls, no external side effects; reads local Move files and reports findings, which is least privilege. Auto-fix is only offered interactively, never forced. Deducted for missing explicit statements on read scope, data flow, rollback, and confirmation norms for the auto-fix phase.
Rules are self-consistent with a clear phased workflow and concrete report template; but checks rely entirely on model heuristics with no scripts, no tests, and no defined failure-feedback path for abnormal input (missing Move.toml, empty directories). Deducted for undefined failure modes; capped by static review.
Trigger conditions (.move files, Move.toml, 'check Move code quality') are semantically clear; but capability boundaries and non-fit ranges are undeclared, rules are tightly bound to Sui 1.45+/Move 2024, there is no Chinese-language support, and referenced resources (Move Book) are overseas-hosted with no mainland-China reachability assessment — hence deductions.
README, CONTRIBUTING and MIT LICENSE present with good documentation structure and official source attribution; but the checklist/, examples/ and tests/ directories described in CONTRIBUTING are not evidenced, and there is no versioning, changelog, or stated maintenance/update responsibility — hence deductions.
Rule list is concrete (50+ rules, 11 categories) with a directly usable report format and real marginal value over manually consulting the Move Book; but static review cannot verify outputs, and rule currency depends on Sui/Move version updates — deducted for lack of verified representative outputs and comparative-benefit evidence.
Rules are traceable to the official Move Book checklist with clear attribution and reasonable fact/inference separation; but there are no committed tests, no CI execution evidence, and no independent reproduction — capped at a low-to-mid score under static review.
- This is a static source review; nothing was executed and confidence is low.
- Rules are tightly coupled to Sui 1.45+ and Move 2024 Edition; applicability to other Move deployments (e.g. Aptos) is undeclared.
- Reachability of overseas resources (Move Book) from mainland China is unassessed.
- The interactive auto-fix phase lacks mandatory user confirmation and rollback guidance; manually review all suggested changes.
- The checklist/tests structure and version governance described in CONTRIBUTING are not visible in the evidence; rules may lag upstream Move Book updates.
What does this skill do, and when should you use it?
This is a Claude Code skill that embeds Move language code-review expertise into Claude. It checks Move packages against the official Move Book Code Quality Checklist — 11 categories with 50+ specific rules — for Move 2024 Edition best practices. The workflow runs in four phases: discovery, systematic analysis, reporting, and interactive review, producing a severity-tiered structured report. It suits developers building on Sui and other Move ecosystems, whether writing new code or auditing existing packages.
The skill reads your Move.toml manifest (verifying edition is 2024.beta/2024, dependencies are implicit, and named addresses are properly prefixed) and globs all .move files plus test modules with a _tests suffix. It then reviews the code across 11 rule categories: module and use statement syntax, constant naming (EPascalCase error codes / ALL_CAPS constants), struct conventions (Cap suffix, past-tense events), function signatures (public vs. entry, parameter ordering, getter naming), body idioms (method chaining, macros, .. unpack syntax), Option and loop macros, testing style (merged #[test, expected_failure], assert_eq!, avoiding unnecessary TestScenario), and doc comments. It outputs a Markdown report with pass/recommendation/critical counts, file paths and line numbers, good/bad code examples, and fix snippets, then optionally answers follow-ups or proposes fixes.
- A Sui smart-contract developer wants Claude to review a module against Move 2024 conventions before opening a PR
- A team auditing an existing Move package wants legacy patterns (curly-brace modules, explicit framework deps) flagged category by category
- A beginner has just written a Move.toml and wants the edition, dependency, and named-address configuration validated
- A developer wants feedback on test style — attributes, naming, and assertion habits — beyond what the compiler reports
- A code reviewer wants to verify a function's parameter ordering, public/entry usage, and getter naming before commenting
What are this skill's strengths and limitations?
- Rules come directly from the official Move Book Code Quality Checklist and are cited in reports
- All 50+ rules include good/bad code examples; reports carry file paths, line numbers, and copyable fix snippets
- Findings are severity-tiered (missing Move 2024 Edition is critical), so fixes can be prioritized
- Pure prompt-based skill — no extra CLI or scripts to install, works right after cloning
- Supports interactive follow-up questions and automatic fix proposals after the report
- Static, prompt-based review only — it does not run the Move compiler or formatter, so it cannot guarantee full violation coverage
- Rules target Move 2024 Edition and Sui 1.45+; applicability to older versions or other Move chains (e.g., Aptos) is unstated
- The repository has no topics, no test suite, and no evidence of real-world results beyond its own documentation
- Formatting requirements (e.g., blank lines between report fields) rely on model compliance, so output layout may vary
How do you install this skill?
Clone the repository into the Claude Code skills directory:
bash
git clone https://github.com/1NickPappas/move-code-quality-skill ~/.claude/skills/move-code-quality
If the directory doesn't exist, run mkdir -p ~/.claude/skills first. The skill lives in the repo's SKILL.md. The source documents no other installation channels (e.g., a plugin marketplace).
How do you use this skill?
The skill activates automatically when you work in a directory containing .move files or Move.toml; you can also prompt explicitly, e.g. "Analyze this Move package for code quality issues", "Review this module against the Move code quality checklist", or "Check if this code follows Move 2024 best practices". At analysis start it asks whether you want a full package scan or a specific file/category. Runtime and token cost are not documented.