ai-generated-business-code-review

AI Business Code Review

Overview

Review AI-generated business/application code for correctness, robustness, maintainability, performance, and security. Output a 0-10 score, a risk level, and a must-fix checklist. For C++ code, REQUIRED: use openharmony-cpp and openharmony-security-review as hard constraints.

When to Use

• AI-generated business/app code review or quality evaluation
• Need scoring, risk level, or must-fix checklist
• C++ business code must meet OpenHarmony coding/security requirements

Workflow

1. Identify language and code type (business vs test). If test code, use ai-generated-ut-code-review instead.
2. For C++: load and apply openharmony-cpp + openharmony-security-review as mandatory constraints.
3. Inspect behavior vs requirements, edge cases, error handling, resource management, and security.
4. Score by rubric, assign risk level, list must-fix items with concrete evidence (file/line or snippet).

Scoring (0-10)

Each dimension 0-2 points. Sum = total score.

Dimension	0	1	2
Correctness	Wrong/missing key behavior	Partial/assumptions	Meets requirements
Robustness	Crashes/leaks/unchecked errors	Some edge handling	Solid edge/error handling
Maintainability	Hard to read/modify	Mixed quality	Clear structure & naming
Performance	Inefficient hot paths	Acceptable but improvable	Efficient for expected load
Security	Clear vulnerabilities	Weak validation/unsafe	Safe-by-default, validated

Risk Levels

• Blocker: security vulnerabilities, severe logic errors, or violations of mandatory C++ specifications
• High: missing critical paths, incomplete exception handling, obvious resource leaks
• Medium: significant maintainability or performance risks
• Low: minor issues or style consistency

Must-Fix Checklist

• External input has explicit validation and error feedback
• Critical exception paths are reachable and observable (logs/error codes/exceptions)
• Resource release or RAII guarantees (C++)
• C++ scenarios satisfy openharmony-cpp and openharmony-security-review

AI-Generated Code Pitfalls (Check Explicitly)

• Misunderstanding of requirements, incorrect implicit assumptions
• Only covers the happy path, ignores exceptions/boundaries
• Swallows exceptions or uses unreasonable default values
• Copy-paste causes duplication and inconsistency
• Overcomplicated or incorrect abstraction

Output Format (Required, Semi-fixed)

• Score: x/10 — Correctness x, Robustness x, Maintainability x, Performance x, Security x
• Risk: Low/Medium/High/Blocker — briefly describe the reason for risk (1 line)
• Must-fix:
  • [Action + Evidence]
  • [Action + Evidence]
• Key Evidence:
  • Cite specific functions/classes/paths (1-2 items)
• Notes:
  • Minimal fix suggestion or alternative approach (1-2 lines)

Rules:

• C++ must be in Key Evidence or Must-fix explicitly point out violations of openharmony-cpp/openharmony-security-review points
• At least 2 pieces of evidence; if evidence is insufficient, explain and deduct points
• Providing only an overall evaluation is prohibited

Common Mistakes

• Only provides an overall evaluation without evidence
• Ignores mandatory C++ specifications
• Treats style issues as blocking risks

Example (Concise)

Score: 6/10 (Correctness 1, Robustness 1, Maintainability 1, Performance 1, Security 2) Risk: High Must-fix:

• Missing input validation in createUser() (null/empty, length, format)
• Error path swallows exceptions without logging Key Evidence:
• createUser() accepts raw input without checks
• catch block is empty in persistUser() Notes:
• Add explicit validation and return structured errors.