Code review with confidence-based filtering. Modes: quick (lint+types), standard (Opus), thorough (parallel agents). Auto-fixes >=95% confidence issues.
/plugin marketplace add GGPrompts/TabzChrome/plugin install conductor@tabz-chromeThis skill inherits all available tools. When active, it can use any tool Claude has access to.
Automated code review with confidence-based filtering and test coverage assessment.
/conductor:code-review # Standard review (Opus)
/conductor:code-review --quick # Fast: lint + types + secrets only
/conductor:code-review --thorough # Deep: parallel specialized reviewers
/conductor:code-review <issue-id> # Review for specific issue
Reviews code with precision scoring to minimize false positives:
| Score | Meaning | Action |
|---|---|---|
| 0 | False positive / pre-existing | Skip |
| 25 | Might be real, can't verify | Skip |
| 50 | Real but minor nitpick | Skip |
| 75 | Likely real but uncertain | Skip |
| 80-94 | Verified issue | Flag |
| 95-100 | Certain bug or rule violation | Auto-fix |
New: Every review now includes a needs_tests assessment.
The reviewer evaluates whether changes warrant test coverage based on:
| Factor | Indicators |
|---|---|
| Complexity | Cyclomatic complexity, lines changed, branching logic |
| Risk Areas | Auth, payments, data mutations, API changes |
| Missing Coverage | New functions/classes without corresponding tests |
| Regression Risk | Bug fixes that could recur, edge cases discovered |
{
"needs_tests": true,
"test_assessment": {
"recommendation": "required",
"rationale": "New API endpoint with input validation and error handling",
"suggested_tests": [
{
"type": "unit",
"target": "validateUserInput()",
"cases": ["valid input", "empty input", "malformed input"]
},
{
"type": "integration",
"target": "POST /api/users",
"cases": ["success path", "validation errors", "auth failure"]
}
],
"priority": "high",
"auto_writable": false
}
}
| Level | When | Action |
|---|---|---|
required | New logic, complex branching, risk area | Block until tests added |
recommended | Moderate changes, some complexity | Flag for consideration |
optional | Simple changes, low risk | Note but don't block |
skip | Docs, config, formatting, existing test coverage | No tests needed |
For issues with >=95% confidence:
Safe to auto-fix: Unused imports, console.log statements, formatting, typos
Never auto-fix: Logic changes, security issues, test coverage gaps
--quick)Fast checks for trivial changes:
skip (trivial changes)Spawns conductor:code-reviewer agent (Opus):
--thorough)Parallel specialized reviewers:
{
"passed": true,
"mode": "standard",
"summary": "Reviewed 5 files. Auto-fixed 2 issues. No blockers.",
"claude_md_checked": ["CLAUDE.md"],
"auto_fixed": [
{"file": "src/utils.ts", "line": 45, "issue": "Unused import", "confidence": 98}
],
"flagged": [
{"severity": "important", "file": "src/api.ts", "line": 23, "issue": "Missing error handling", "confidence": 85}
],
"blockers": [],
"needs_tests": true,
"test_assessment": {
"recommendation": "recommended",
"rationale": "New utility function with multiple code paths",
"suggested_tests": [
{"type": "unit", "target": "formatDate()", "cases": ["valid date", "invalid date", "null input"]}
],
"priority": "medium",
"auto_writable": true
}
}
The /conductor:worker-done pipeline uses test assessment:
needs_tests: true and recommendation: required:
auto_writable: true:
When auto_writable: true, can spawn test-writer:
Task(
subagent_type="general-purpose",
prompt="Write tests for changes. Assessment: ${test_assessment}"
)
recommendation: required)| Resource | Purpose |
|---|---|
plugins/conductor/agents/code-reviewer.md | Agent implementation |
plugins/conductor/commands/code-review.md | Command definition |
/conductor:worker-done | Full completion pipeline |
/conductor:run-tests | Test execution |
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.