Code Review

Tepora Code Review Skill

Skill Purpose

This skill enables the agent to perform comprehensive code reviews for the Tepora project, a privacy-focused local LLM assistant with episodic memory (EM-LLM) built with Rust, Tauri, React, and TypeScript.

When to Use This Skill

Trigger this skill when:

• User asks to "review this code" or "check this PR" for Tepora project
• User requests quality assessment of Tepora codebase
• User wants architectural analysis of Tepora components
• User needs security or performance review of Tepora code
• User mentions "code review", "CR", "analyze code", or "check implementation" in Tepora context

Core Principles

1. Context-Aware Review

Understand Tepora's unique characteristics:

• Privacy-First: All data stays local, no cloud uploads
• Memory System: EM-LLM for episodic memory storage
• Dual-Agent: Main agent (Tepora) + Professional agent
• Cyber Tea Salon: Focus on beautiful, calming UX
• Tech Stack: Rust (Tauri backend) + React/TypeScript (frontend)

2. Multi-Layered Analysis

Review must cover:

• Code Quality (syntax, patterns, maintainability)
• Architecture Alignment (fits Tepora's design)
• Security & Privacy (data protection, no leaks)
• Performance (memory, CPU, startup time)
• User Experience (UI/UX quality)

Review Process

Step 1: Initialize Review Context

# Before starting, understand project structure
# Use list_dir and view_file to explore
ls -R /path/to/tepora-project
cat /path/to/tepora-project/README.md
cat /path/to/tepora-project/docs/architecture/ARCHITECTURE.md

Output: Brief summary of project state and review scope.

Step 2: Identify Review Scope

Determine what to review:

For PR Review:

# If reviewing a specific PR or branch
git diff --name-only main...feature-branch
git diff main...feature-branch

For Module Review: Focus on specific directories:

• Tepora-app/src-tauri/ - Rust backend
• Tepora-app/src/ - React/TS frontend
• docs/ - Documentation
• .github/ - CI/CD workflows

For Full Project Review: Review all major components systematically.

Step 3: Execute Detailed Analysis

For each file reviewed, apply the appropriate checklist:

Rust Code Review (src-tauri/)

# Run automated checks first
cd Tepora-app && cargo clippy --all-features -- -D warnings
cd Tepora-app && cargo fmt -- --check
cd Tepora-app && cargo test

Manual Review Checklist:

1. Ownership & Lifetimes

   • Are borrows and moves handled correctly?
   • Are lifetimes explicit where needed?
   • Is there unnecessary cloning?

2. Error Handling

   • Are Result and Option used appropriately?
   • Are errors propagated with ? operator?
   • Are custom error types defined for domain errors?
   • Are panics avoided in production code?

3. Concurrency & Safety

   • Are Arc and Mutex used correctly?
   • Are race conditions prevented?
   • Is unsafe code justified and documented?
   • Are thread boundaries clear?

4. Tauri Integration

   • Are commands properly annotated with #[tauri::command]?
   • Is state management handled via Tauri's state system?
   • Are IPC calls efficient and minimal?

5. EM-LLM Specifics

   • Is memory storage efficient (serialization, compression)?
   • Are memory queries optimized (indexing, caching)?
   • Is privacy maintained (encryption, no logging of sensitive data)?

6. Performance

   • Are allocations minimized in hot paths?
   • Are data structures optimal (Vec vs HashMap vs BTreeMap)?
   • Is lazy evaluation used where appropriate?

Example Analysis Output:

### File: src-tauri/src/memory/episodic.rs

**Strengths:**
- ✅ Good use of `serde` for serialization
- ✅ Proper error handling with custom `MemoryError` type
- ✅ Thread-safe access with `Arc<RwLock<T>>`

**Issues:**
- ⚠️ Line 45: Unnecessary `.clone()` on `episode_id` - can use reference
- ⚠️ Line 78: Consider using `BTreeMap` instead of `HashMap` for sorted keys
- ❌ Line 120: Sensitive data (user content) stored without encryption

**Recommendations:**
1. Add encryption layer for stored memories
2. Implement LRU cache for frequently accessed memories
3. Add comprehensive unit tests for edge cases

TypeScript/React Code Review (src/)

# Run automated checks
cd Tepora-app && npm run type-check
cd Tepora-app && npm run lint
cd Tepora-app && npm test

Manual Review Checklist:

1. Type Safety

   • Are types explicit (avoid any)?
   • Are interfaces/types properly defined?
   • Are generics used effectively?

2. React Best Practices

   • Are hooks used correctly (dependencies array)?
   • Is component re-rendering optimized (memo, useMemo, useCallback)?
   • Are side effects properly managed (useEffect)?
   • Is state lifted only when necessary?

3. Component Design

   • Is separation of concerns clear (UI vs logic)?
   • Are components small and focused?
   • Is composition preferred over inheritance?

4. Accessibility

   • Are ARIA labels present?
   • Is keyboard navigation supported?
   • Is color contrast sufficient?

5. Cyber Tea Salon UX

   • Is design consistent with the "tea salon" aesthetic?
   • Are animations smooth (60fps)?
   • Are loading states handled gracefully?
   • Are error messages user-friendly?

Example Analysis Output:

### File: src/components/ChatInterface.tsx

**Strengths:**
- ✅ Good separation of chat logic into custom hook
- ✅ Accessible with proper ARIA labels
- ✅ Smooth scroll-to-bottom animation

**Issues:**
- ⚠️ Line 34: Missing dependency in useEffect array (`userId`)
- ⚠️ Line 67: Inefficient - re-creating `handleSubmit` on every render
- ❌ Line 89: Using `any` type for message object

**Recommendations:**
1. Add `userId` to useEffect dependencies or use useCallback
2. Wrap `handleSubmit` in `useCallback` with proper dependencies
3. Define explicit `Message` interface
4. Consider virtualization for long chat histories

Security Review

Always check:

1. Data Privacy

   • No data sent to external servers without consent
   • Sensitive data encrypted at rest
   • Secure key management

2. Input Validation

   • User inputs sanitized
   • SQL/NoSQL injection prevented
   • XSS vulnerabilities avoided

3. Dependencies

   cargo audit
   npm audit
   

4. Secrets Management

   grep -r "password\|secret\|api_key" --include="*.rs" --include="*.ts" src/
   

Step 4: Performance Analysis

# Rust performance
cargo build --release
cargo bench  # if benchmarks exist

# Frontend bundle analysis
npm run build
npx vite-bundle-visualizer

Check:

• Startup time < 3 seconds
• Memory usage reasonable (< 500MB for typical use)
• Bundle size optimized (< 5MB)
• CPU usage not excessive

Step 5: Generate Review Report

Create a comprehensive report using this template:

# Tepora Code Review Report

**Date**: [Current Date]
**Reviewer**: AI Agent
**Scope**: [PR #XXX / Module / Full Project]
**Status**: [✅ Approved / ⚠️ Minor Changes / ❌ Major Changes]

---

## Executive Summary

[2-3 sentence overview of findings]

## Strengths 🎯

1. [Positive finding 1]
2. [Positive finding 2]
3. ...

## Critical Issues ❌

### Issue 1: [Title]
- **File**: `path/to/file.rs:123`
- **Severity**: Critical
- **Description**: [What's wrong]
- **Impact**: [Why it matters]
- **Fix**: [How to resolve]
- **Code Example**:
  ```rust
  // Before
  [problematic code]
  
  // After
  [suggested fix]

Major Issues ⚠️

[Same format as Critical]

Minor Issues 💡

[Same format, but less urgent]

Metrics 📊

• Test Coverage: XX%
• Clippy Warnings: XX
• ESLint Warnings: XX
• Build Time: XX seconds
• Bundle Size: XX KB
• Security Vulnerabilities: XX

Architecture Analysis 🏗️

[Discussion of architectural patterns, consistency, scalability]

Security Analysis 🔒

[Privacy, data protection, vulnerability assessment]

Performance Analysis ⚡

[Speed, memory, resource usage analysis]

Recommendations 💡

Short-term (This Sprint)

1. [Action item]
2. [Action item]

Long-term (Future Iterations)

1. [Improvement idea]
2. [Improvement idea]

Conclusion

[Summary paragraph with overall assessment]

Next Steps: [What should happen next]

## Advanced Review Techniques

### 1. Architectural Pattern Recognition

Identify and validate patterns:
- **Command Pattern**: Tauri commands for frontend-backend communication
- **Repository Pattern**: EM-LLM memory storage abstraction
- **Observer Pattern**: React state management
- **Factory Pattern**: Agent creation and management

### 2. Privacy-First Validation

For Tepora specifically, always verify:
```rust
// ✅ Good: Local storage
let memory_path = app_data_dir().join("memories.db");

// ❌ Bad: External API call without user consent
let response = reqwest::get("https://api.example.com/log")
    .await?;

3. Memory Efficiency Check

Look for memory-intensive operations:

// ⚠️ Potential issue: Loading entire history into memory
let all_memories: Vec<Episode> = db.get_all_episodes()?;

// ✅ Better: Pagination or streaming
let recent_memories = db.get_episodes_paginated(0, 50)?;

4. Concurrency Analysis

Verify thread safety:

// ✅ Good: Proper synchronization
let state = Arc::new(RwLock::new(AppState::new()));

// ❌ Bad: Shared mutable state without synchronization
static mut GLOBAL_COUNTER: i32 = 0;

Common Pitfalls to Watch For

Rust-Specific

1. Unnecessary Cloning

   // ⚠️ Bad
   fn process(data: Vec<String>) {
       for item in data.clone() { ... }
   }
   
   // ✅ Good
   fn process(data: &[String]) {
       for item in data { ... }
   }
   

2. Blocking in Async Context

   // ❌ Bad
   async fn load_memory() {
       std::thread::sleep(Duration::from_secs(1)); // blocks executor
   }
   
   // ✅ Good
   async fn load_memory() {
       tokio::time::sleep(Duration::from_secs(1)).await;
   }
   

TypeScript/React-Specific

1. Infinite Re-render Loop

   // ❌ Bad
   useEffect(() => {
       setState({ ...state, updated: true });
   }); // Missing dependency array!
   
   // ✅ Good
   useEffect(() => {
       setState({ ...state, updated: true });
   }, [/* dependencies */]);
   

2. Type Coercion Abuse

   // ⚠️ Bad
   const userId = props.user as any;
   
   // ✅ Good
   interface User { id: string; name: string; }
   const userId: string = props.user.id;
   

Output Format

Always structure your review output as follows:

1. Brief Summary (2-3 sentences)
2. Automated Check Results (if applicable)
3. Detailed Findings (organized by severity)
4. Code Examples (show before/after)
5. Metrics (quantitative data)
6. Actionable Recommendations

Example Review Execution

User: "Review the memory module in src-tauri/src/memory/mod.rs"

Agent Response:

I'll review the memory module following the Tepora code review workflow.

[Runs automated checks]
[Views the file]
[Analyzes code]
[Generates detailed report]

Here's my review:

# Memory Module Review

## Summary
The memory module implements the core EM-LLM functionality. Overall structure is solid, but found 3 critical issues related to encryption and 2 performance optimizations.

[... detailed findings ...]

Continuous Improvement

After each review:

1. Learn from patterns found
2. Update checklist if new issues discovered
3. Refine understanding of Tepora's architecture
4. Suggest workflow improvements

Integration with Development Workflow

This skill integrates with:

• Git: For PR reviews and diffs
• CI/CD: Can be triggered in GitHub Actions
• IDE: Via Agent tool integration
• Manual: Direct requests

Limitations

Be aware of:

• Cannot execute code that requires compilation beyond static analysis tools
• Cannot access external APIs for third-party security scans
• May not catch runtime-only issues
• Cannot test UI/UX subjectively (provide heuristic assessment)

Success Criteria

A successful review:

• ✅ Identifies concrete, actionable issues
• ✅ Provides code examples for fixes
• ✅ Considers Tepora's unique context
• ✅ Balances thoroughness with clarity
• ✅ Respects the "Cyber Tea Salon" philosophy

Remember: The goal is not just to find bugs, but to help Tepora become the best "memory-enhanced" AI companion it can be, while maintaining its privacy-first, user-friendly ethos.

Converted and distributed by TomeVault — claim your Tome and manage your conversions.