メニュー
Programmatic code style validation using AST analysis. Complements (not replaces) code-style rules by providing automated checking and instant feedback.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/oimiragieo/agent-studio --skill code-style-validatorコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Agent frontmatter enhancements — disallowedTools, mcpServers scoping, fork_eligible field
Context management — pre-compact persistence, threshold alignment, microcompact/circuit breaker detection
Hook enhancements — updatedInput for bash safety prefixes, suppressOutput for verbose blocks, denial-based routing feedback
SKILL.md を表示中
SOC 職業分類に基づく
| name | code-style-validator |
| description | Programmatic code style validation using AST analysis. Complements (not replaces) code-style rules by providing automated checking and instant feedback. |
| version | 1.1.0 |
| model | haiku |
| invoked_by | both |
| user_invocable | true |
| tools | ["Read","Grep","Bash","Glob"] |
| best_practices | ["Use AST analysis for accurate validation","Provide auto-fix suggestions","Integrate with pre-commit hooks"] |
| error_handling | graceful |
| streaming | supported |
| verified | true |
| lastVerifiedAt | "2026-02-22T00:00:00.000Z" |
| source | builtin |
| trust_score | 100 |
| provenance_sha | 3c302af23964a90d |
No separate download: the skill runs the in-repo tool .claude/tools/cli/security-lint.cjs.
winget install OpenJS.NodeJS.LTS (Windows), brew install node (macOS).AST-based validation: Use ESLint (or equivalent) with selectors/patterns; rules listen for specific node types. Prefer existing community rules before writing custom ones; custom rules need meta (type, fixable, schema) and create.
Process: Check naming (camelCase/PascalCase/snake_case), indentation, import order, function/class structure, comment style. Run in pre-commit and CI; provide auto-fix where possible.
Hacks: Use --fix for auto-fixable rules; combine with security-lint for policy. Define rule metadata clearly for docs and tooling. Use AST selectors for precise pattern matching (node type, attributes, child/sibling).
ESLint: Use in project, custom rules, selectors. No cert; skill data: AST-based validation, pre-commit/CI, auto-fix, naming/indent/imports.
Suggested hooks: Pre-commit: run security-lint (skill script) or ESLint; block commit on failure. Use with developer (secondary), code-reviewer (secondary), qa (CI).
Workflows: Use with code-reviewer (secondary), developer (secondary). Flow: before commit or in CI → run validator → fix or block. See code-review-workflow.md, validation hooks.
Analyze the codebase to identify style patterns:
Use language-specific AST parsers:
TypeScript/JavaScript: </execution_process>
const ts = require('typescript');
const fs = require('fs');
function validateTypeScriptFile(filePath) {
const sourceCode = fs.readFileSync(filePath, 'utf8');
const sourceFile = ts.createSourceFile(filePath, sourceCode, ts.ScriptTarget.Latest, true);
const issues = [];
// Validate naming conventions
ts.forEachChild(sourceFile, node => {
if (ts.isFunctionDeclaration(node) && node.name) {
if (!/^[a-z][a-zA-Z0-9]*$/.test(node.name.text)) {
issues.push({
line: sourceFile.getLineAndCharacterOfPosition(node.name.getStart()).line + 1,
message: `Function name "${node.name.text}" should be camelCase`,
});
}
}
});
return issues;
}
</code_example>
<code_example> Python AST Validation:
import ast
import re
def validate_python_file(file_path):
with open(file_path, 'r') as f:
source_code = f.read()
tree = ast.parse(source_code)
issues = []
for node in ast.walk(tree):
if isinstance(node, ast.FunctionDef):
if not re.match(r'^[a-z][a-z0-9_]*$', node.name):
issues.append({
'line': node.lineno,
'message': f'Function name "{node.name}" should be snake_case'
})
return issues
</code_example>
<code_example> Style Checks:
Naming Conventions:
Formatting:
Structure:
<code_example> Usage Examples (Template - implement validator first):
Validate Single File:
# After implementing your validator:
node validate-code-style.js src/components/Button.tsx
Validate Directory:
# After implementing your validator:
node validate-code-style.js src/components/
Output Format:
{
"file": "src/components/Button.tsx",
"valid": false,
"issues": [
{
"line": 15,
"column": 10,
"rule": "naming-convention",
"message": "Variable 'UserData' should be camelCase: 'userData'",
"severity": "error"
},
{
"line": 23,
"column": 5,
"rule": "indentation",
"message": "Expected 2 spaces, found 4",
"severity": "warning"
}
],
"summary": {
"total": 2,
"errors": 1,
"warnings": 1
}
}
</code_example>
<code_example> Pre-commit Hook (Template - implement validator first):
#!/bin/bash
# .git/hooks/pre-commit
changed_files=$(git diff --cached --name-only --diff-filter=ACM | grep -E '\.(ts|tsx|js|jsx|py)$')
for file in $changed_files; do
# Replace with your actual validator path
if ! node validate-code-style.js "$file"; then
echo "Code style validation failed for $file"
exit 1
fi
done
</code_example>
<code_example> CI/CD Integration (Template - implement validator first):
# .github/workflows/code-style.yml
- name: Validate code style
run: |
# Replace with your actual validator path
node validate-code-style.js src/
if [ $? -ne 0 ]; then
echo "Code style validation failed"
exit 1
fi
</code_example>
| Anti-Pattern | Why It Fails | Correct Approach |
|---|---|---|
| Regex-based style checking | Breaks on multi-line, strings, and comments | Use AST-based tools (ESLint, Prettier) |
| Custom rules for covered standards | High maintenance; inconsistent with community | Configure existing ESLint/Prettier rules first |
| Blocking CI on warnings | Excessive friction causes teams to disable style checks | Block only on errors; log warnings |
| Style validation without auto-fix | Manual fixes have low resolution rate | Always provide auto-fix commands alongside findings |
| Running only locally, not in CI | Developers bypass pre-commit hooks regularly | Run in both pre-commit and CI pipeline |
Before starting:
Read .claude/context/memory/learnings.md
After completing:
.claude/context/memory/learnings.md.claude/context/memory/issues.md.claude/context/memory/decisions.mdASSUME INTERRUPTION: If it's not in memory, it didn't happen.