17th-test-analyzer

name: 17th-test-analyzer description: | Use this skill when user asks about test coverage, test sufficiency, or whether tests cover requirements. Analyzes staged changes against task spec using GPT-5.3-Codex to find gaps in happy cases and edge cases. Auto-triggers on: "테스트 충분해?", "test coverage", "edge case", "테스트 분석" Appropriate for: test gap analysis, coverage check, pre-PR test review. Not appropriate for: running tests, writing tests, code review.

Test Analyzer

Analyze test coverage for staged changes against task specification using GPT-5.3-Codex.

Usage

/test-analyzer                    → Codex high, TASK.md (default)
/test-analyzer deep               → Codex xhigh, TASK.md
/test-analyzer TODO.md            → Codex high, TODO.md
/test-analyzer TODO.md deep       → Codex xhigh, TODO.md

Argument Parsing

Keywords (reserved):

• deep → Use xhigh effort instead of high

Other arguments → Task spec file name

Parsing Logic

args = split(input)
task_file = "TASK.md"
effort = "high"

for arg in args:
    if arg == "deep":
        effort = "xhigh"
    else:
        task_file = arg

Result Mapping

Implementation

Invoke the 17th-test-analyzer subagent:

Task(
    description="Analyze test coverage ({effort}, {task_file})",
    subagent_type="17th-test-analyzer",
    prompt="""
Analyze test coverage for staged changes.

Configuration:
- Task file: {task_file}
- Effort: {effort}  (high or xhigh)
- Project root: {project_root}
""",
    run_in_background=true
)

Auto-Trigger

When skill auto-triggers (no args), uses defaults:

• Effort: high
• Task file: TASK.md

Output

• Report: .kkachi/{branch_name}/completed/test-analyzer-{epoch_timestamp}.md
• Summary: X/Y requirements covered, Z gaps found
• Recommendations: prioritized action items

Examples

Notes

• Uses GPT-5.3-Codex for all analysis (no GLM)
• Analyzes staged changes only
• Supports Go, Python, Dart, TypeScript, JavaScript
• Results verified before reporting
• Ensure .kkachi/ is in .gitignore