Install
$ agentstack add skill-leekjay-claude-skills-plugin-code-quality-standards ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Code Quality Standards
Ensure code quality and project maintainability through standardized check workflows.
Decision Framework
Default behavior: For comprehensive code quality checks, prefer Sub-Agent mode to ensure thorough validation.
🔴 Mandatory Sub-Agent Triggers (ANY ONE triggers delegation)
- Full quality check: User requests complete code quality validation before commit/PR
- Multiple check types: Needs format + lint + typecheck + test (2+ check types)
- Fix and verify cycle: Errors found that need fixing and re-verification
- Large changeset: Changes span 5+ files or significant code modifications
- Pre-release check: Final validation before release or deployment
- CI failure investigation: Investigating why CI checks failed
- Type error fixing: Multiple type errors that need systematic fixing
- Iterative fixing: Fixes may introduce new errors requiring multiple rounds
🟢 Main Conversation Handling (ALL conditions must be met)
- Single check: Only running one command (e.g., "run lint" or "run tests")
- Quick verification: Just checking if something passes, not fixing
- Known simple fix: Already know what to fix, just need to apply it
- No iteration needed: Single pass expected to succeed
Decision Flow
Check for ANY 🔴 mandatory trigger?
├─ YES → ✅ USE SUB-AGENT MODE immediately
│ subagent_type="code-quality-standards:code-quality-checker"
└─ NO → Check if ALL 🟢 simple conditions are met?
├─ YES → Handle in main conversation
└─ NO → ✅ USE SUB-AGENT MODE (default behavior)
Quick Reference Examples
| User Description | Trigger Signal | Decision | |-----------------|----------------|----------| | "Run all quality checks" | 🔴 Full quality check | Sub-Agent | | "Just run pnpm lint" | 🟢 Single check | Main conversation | | "Fix all type errors" | 🔴 Type error fixing | Sub-Agent | | "Check if tests pass" | 🟢 Quick verification | Main conversation | | "Prepare code for PR" | 🔴 Pre-release check | Sub-Agent | | "Run format" | 🟢 Single check | Main conversation | | "Why is CI failing?" | 🔴 CI failure investigation | Sub-Agent | | "I changed 10 files, please verify" | 🔴 Large changeset | Sub-Agent |
When to Execute
✅ Must execute full checks:
- After completing feature implementation or bug fixes
- When user explicitly requests code quality checks
- Before creating Git commits or PRs
❌ Not needed:
- Minor text edits or formatting adjustments
- Documentation-only changes
- Exploratory code reading
Check Workflow (Execute in Order)
1. Code Formatting
pnpm format
Ensure all files comply with project code style standards.
2. Code Linting
pnpm lint
- Fix all errors and warnings
- Don't use
eslint-disableto bypass issues
3. Type Checking (TypeScript)
pnpm typecheck
# If typecheck script doesn't exist, use:
tsc --noEmit
Type Error Fixing Rules (Only when fixing type errors)
❌ Absolutely Prohibited
- Type assertions:
as,as any,as unknown as X anytype declarations@ts-ignoreor@ts-expect-errorcomments
✅ Mandatory Approach
- Analyze root cause - Understand why the type error occurs
- Correct actual types - Fix type definitions, declare types explicitly
- Improve type definitions - Specify parameter and return types explicitly
- Add type guards - Use type guards for TypeScript auto-inference
- Refactor code structure - If type errors are hard to fix, may be a design issue
📝 Rationale
- Type assertions and any hide real problems, leading to runtime errors
- Loses type safety and IDE intellisense
- Accumulates technical debt
- Misleads other developers
⚠️ Exception Scenarios (Must consult user first)
- Third-party library with incomplete type definitions
- Library lacks proper type definitions
- Consult user first, consider contributing type definitions or using alternative library
- Complex dynamic type scenarios
- Complex dynamic types where proper typing requires significant architectural changes
- Consult user first, evaluate refactoring cost vs technical debt
- TypeScript type system limitations
- Legitimate edge cases that TypeScript's type system cannot properly express
- Consult user first, confirm this is indeed a TypeScript limitation
- Add detailed comments explaining why type assertion is needed
🔴 Complex Type Errors
If type error is complex or difficult to understand:
- Stop immediately - Don't continue trying to fix
- Consult user - Seek guidance
- Don't use type assertions or any to bypass
📌 Important Note
These strict rules only apply when fixing type errors!
For general code writing:
- Can follow existing project patterns
- No need to refactor all existing type assertions
- Follow existing code style
4. Unit Tests (if applicable)
pnpm test
Ensure all tests pass.
Failure Handling Workflow
- Fix issues immediately - Analyze error cause, apply correct fix
- Re-run failed check - Confirm issue is resolved
- Re-run all checks - If fix introduces new code changes, restart from step 1
- Keep todo status - Until all checks pass
- Don't commit failing code
Completion Confirmation
After all checks pass, report:
✅ All code quality checks passed
- ✓ Code formatting (pnpm format)
- ✓ Code linting (pnpm lint)
- ✓ Type checking (pnpm typecheck)
- ✓ Unit tests (pnpm test)
Code is ready to commit.
Special Cases
Check Commands Don't Exist
- Check
package.jsonscriptssection - Look for alternative commands
- If truly doesn't exist, notify user and suggest configuration
Multiple Checks Fail Simultaneously
Fix in order: format → lint → typecheck → test
Check Runs Over 5 Minutes
Notify user and ask whether to continue waiting or interrupt to investigate.
Unable to Fix Check Errors
Stop fix attempts, consult user, provide error details and attempted fixes.
Recommended Tools
Git Hooks
{
"husky": {
"hooks": {
"pre-commit": "pnpm lint-staged",
"pre-push": "pnpm typecheck && pnpm test"
}
}
}
CI/CD Automated Checks
Run all checks automatically in CI/CD pipeline to ensure every commit and PR passes quality checks.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: LeekJay
- Source: LeekJay/claude-skills-plugin
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.