- Home
- Skills
- Code Quality & Review
- Tool-AI Code Reviewer
Works with the AI Agents you already use
Tool-AI Code Reviewer
Unlike generic code review prompts, this skill follows a structured, multi-pass review process and requires evidence for every reported finding.
$1.45
Tool-AI Code Reviewer
Example session with this skill installed
Review this Python snippet for security and performance issues:
def get_user(request):
user_id = request.GET.get('id')
query = "SELECT * FROM users WHERE id = " + user_id
return db.execute(query)
- Read your context and instructions
- Compiled the tool-ai code reviewer
Code Review Report
Executive Summary
Static review of one Python function, get_user (4 lines). No code was executed. The function builds a SQL query by concatenating a request parameter directly into the query string, which is a confirmed SQL injection. It also fetches every column with SELECT * and has no handling for a missing id parameter. The code is not safe to deploy as written.
Overall Risk Level
Critical — user-controlled input reaches a SQL query unescaped.
Critical Findings
[C-1] SQL injection in get_user
- File / Location:
get_user, the line buildingquery - Severity: Critical
- Confidence: Confirmed
Explanation: user_id comes from request.GET.get('id') and is concatenated into the SQL string. A value such as 1 OR 1=1 returns every row, and 1; DROP TABLE users may run a second statement depending on the driver.
Why it matters: An attacker can read other users' records, bypass lookups, or modify data, depending on database permissions. The source is a public query-string parameter, so no authentication is needed to reach it.
- Suggested fix: Pass the value as a bound parameter instead of building the string.
Security Findings
No additional non-critical security findings. Authorization (whether the caller may read this user) could not be assessed because the caller was not provided.
Performance Findings
[P-1] SELECT * fetches unused columns
- File / Location:
get_user, thequerystring - Severity: Low
- Confidence: Confirmed
- Explanation: All columns are fetched, including any large or sensitive ones such as password hashes.
Why it matters: Extra data is transferred on every call, and sensitive columns may be exposed to code that does not need them.
- Suggested fix: Select only the needed columns.
Code Quality Findings
[Q-1] Missing or non-numeric id is not handled
- File / Location:
get_user, first line - Severity: Medium
- Confidence: Confirmed
Explanation: If id is absent, user_id is None and the string concatenation raises a TypeError, which surfaces as a server error.
- Why it matters: A malformed request crashes the handler instead of returning a clean 400 or 404.
- Suggested fix: Validate and convert the value before querying.
Suggested Fixes
- Replace string concatenation with a parameterized query (C-1).
- Validate that
idexists and is an integer (Q-1). - Select explicit columns instead of
*(P-1).
Improved Code Examples
def get_user(request):
raw_id = request.GET.get("id")
if raw_id is None or not raw_id.isdigit():
return None # or return a 400 response in your framework
return db.execute(
"SELECT id, username, email FROM users WHERE id = %s",
(int(raw_id),),
)
The placeholder style depends on your driver (%s for psycopg2 and MySQL drivers, ? for sqlite3).
Review Limitations
Static review only; nothing was executed or tested. The db object, the calling view, and the database schema were not provided, so the placeholder syntax and the authorization model are marked as needing verification.
Connects securely to your tools. The creator never sees your data.
About this skill
DevKit Pro – AI Code Reviewer
Get reliable, evidence-based code reviews without the noise of generic AI feedback.
DevKit Pro is an AI-powered code review skill designed to help developers identify real bugs, security vulnerabilities, performance bottlenecks, and maintainability issues before they become production problems.
Unlike generic code review prompts, this skill follows a structured, multi-pass review process and requires evidence for every reported finding.
What It Does
- Multi-language Support: Reviews JavaScript, TypeScript, Python, Java, PHP, Go, Rust, and C#.
- Security Analysis: Identifies vulnerabilities, trust boundaries, unsafe input handling, and dangerous data flows.
- Bug Detection: Finds logic errors, edge cases, incorrect assumptions, and potential runtime failures.
- Performance Analysis: Detects inefficient algorithms, unnecessary operations, repeated I/O, and potential bottlenecks.
- Code Quality: Evaluates maintainability, readability, error handling, and architectural concerns.
- Evidence-Based Findings: Requires specific code references and supporting evidence for reported issues.
- Risk Prioritization: Classifies findings by severity: Critical, High, Medium, and Low.
How It Works
- Understand the Context: Identifies the programming language, framework, and purpose of the code.
- Inspect the Code: Performs separate passes for correctness, security, performance, and maintainability.
- Verify Findings: Checks each potential issue against the actual source code to reduce false positives.
- Generate a Report: Produces a structured review with severity levels, explanations, and actionable fixes.
Supported Languages
JavaScript, TypeScript, Python, Java, PHP, Go, Rust, and C#.
Why Use DevKit Pro?
Generic AI prompts can produce noisy feedback, overlook important context, or report problems that do not exist.
DevKit Pro uses a consistent review methodology that emphasizes:
- Real issues over stylistic nitpicking.
- Evidence over assumptions.
- Actionable fixes over vague recommendations.
- Security and correctness over superficial feedback.
Use Cases
- Reviewing code before merging pull requests.
- Auditing authentication and authorization logic.
- Identifying SQL injection and other security risks.
- Finding performance bottlenecks.
- Improving legacy code.
- Reviewing code changes before production deployment.
Limitations
DevKit Pro performs static code analysis. It does not execute, compile, or run tests against the reviewed code.
Issues that depend on unavailable configuration, external services, or database behavior are clearly identified as requiring further verification.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Code Quality & Review.
Static review only. Cannot execute, compile, or profile code. Findings depending on unseen external context are marked as needing verification. Line numbers require provided context.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 16 days ago
Permissions
File Scopes
Read-only skill. It reads the source files you point it at and writes its report in the chat. It does not execute code, modify files, or make network requests.