Skip to main content
Unlisted page
This page is unlisted. Search engines will not index it, and only users having a direct link can access it.

Archived Skill: Code Analysis and Optimization

Skill Name​

code-analysis-and-optimization

Description​

A SHAFT-focused analysis skill for identifying reliability and performance issues, ranking root causes, and proposing minimal-risk fixes with clear validation steps.

High-Value Enhancements for Accuracy​

1) Root-Cause-First Diagnosis​

  • Distinguish symptom, trigger, and root cause for each issue.
  • Require code-location evidence before assigning high confidence.
  • Separate primary cause from secondary suspects.

2) Confidence Scoring​

  • High: direct anti-pattern match in provided code.
  • Medium: partial evidence with probable mapping.
  • Low: missing critical files/flows.

3) Fix Ordering by Risk​

  • Prioritize by impact and blast radius: leaks/flakiness first, then maintainability.
  • Recommend minimal viable fix first, then optional improvements.

4) Validation-Centric Output​

  • Every suggested fix includes explicit “how to verify” checks.
  • Include regression checks for adjacent code paths.

SHAFT-Critical Checks​

  1. ThreadLocal cleanup requires both quit() and remove().
  2. @AfterMethod(alwaysRun = true) for reliable cleanup.
  3. Prefer SHAFT assertion API over raw test framework asserts.
  4. Reused locators should be cached as class-level static final values.
  5. Global SHAFT.Properties.flags mutations must be isolated and restored.

Structured Output​

## SHAFT Code Analysis Report

### Summary
- Total issues found: [count]
- Critical: [n] | High: [n] | Medium: [n] | Low: [n]
- Confidence: High / Medium / Low

### Recommended Fix Order
1. ...

### Issue #[N]
| Field | Value |
|---|---|
| Severity | Critical / High / Medium / Low |
| Confidence | High / Medium / Low |
| Category | ThreadLocal / WebDriver / Assertion / Locator / Concurrency / Performance / TestDesign |
| Location | `path/File.java:line` |
| Description | Root-cause explanation |
| Impact | Runtime or maintenance impact |
| Fix | Minimal corrected snippet |
| Validation | Exact checks/tests to run |

Trigger Examples​

  • “Find why this SHAFT test is flaky and propose a minimal fix.”
  • “Analyze this file for ThreadLocal and teardown leaks.”
  • “Rank the top fixes needed for this failing SHAFT test class.”

Limitations​

  • Static-analysis driven; runtime profilers are not executed automatically.
  • Output quality depends on provided context coverage.
  • Private repositories/files cannot be fetched by the helper script without accessible raw content.