- Home
- Skills
- Code Quality & Review
- podugboch smart code reviewer
podugboch smart code reviewer
Structured automated code reviews with severity ratings and actionable fix suggestions for git diffs and files.
Free
Works with the AI Agents you already use
podugboch smart code reviewer
Example session with this skill installed
Run a high-severity review on my staged changes in the auth module.
- Read your context and instructions
- Compiled the podugboch smart code
I have analyzed your staged changes in the auth module. I found 3 issues meeting the high-severity threshold.
Code Review Report
| # | Severity | Category | File | Location | Issue & Suggestion |
|---|---|---|---|---|---|
| 1 | High | Security | auth/service.py | Line 42 | Hardcoded Secret: The API key is defined as a string constant. Move this to an environment variable using os.getenv("API_KEY"). |
| 2 | High | Bug | auth/utils.py | Line 89 | Unchecked Exception: The token_decode block catches all exceptions but does not log or re-raise, potentially masking invalid signatures. Implement specific JWTError handling. |
| 3 | High | Performance | auth/db.py | Line 112 | N+1 Query: User roles are fetched in a loop after the initial user query. Use a JOIN or selectinload in your SQLAlchemy query to fetch roles in one trip. |
Next steps
- Apply the environment variable fix in
auth/service.pybefore committing. - Add a logging statement to the catch block in
auth/utils.pyto improve observability. - Refactor the database query in
auth/db.pyto use eager loading.
Connects securely to your tools. The creator never sees your data.
About this skill
Manual code reviews are slow, inconsistent, and prone to missing subtle logic flaws or security vulnerabilities. This skill provides a structured, automated audit of your changes to catch critical bugs before they reach production.
What it does
- Multi-scope analysis scans staged changes, branch diffs, single files, or entire directories to find hidden defects.
- Severity grading assigns a clear priority (Critical, High, Medium, Low) to every identified issue.
- Categorized feedback groups findings into security, performance, maintainability, and functional bug categories.
- Actionable suggestions provides concrete code corrections for every flagged line.
- Flexible output generates human-readable text reports or machine-readable JSON for CI/CD integration.
How it works
- Target selection uses arguments like
--staged,--branch, or--fileto define the code surface area. - Context gathering extracts the diff or file content and prepares it for analysis.
- Structured auditing processes the code through an expert review logic to identify specific violations.
- Report generation outputs a numbered list of issues with file locations and fix instructions.
Frameworks & tools
This skill is language-agnostic and works with any text-based source code. It integrates directly with Git for diff management and uses the Anthropic API for reasoning.
Why this beats prompting it yourself
Generic prompts often miss edge cases or provide vague "clean code" advice without context. This skill uses a structured evaluation framework that enforces severity levels and specific categories, ensuring the output is ready for a PR checklist rather than just a chat window.
Use cases
- Audit staged changes before running
git commit. - Perform a final security and performance check before merging a feature branch.
- Scan legacy folders for maintainability and technical debt.
- Generate JSON reports to gate internal development workflows.
Known limitations
Folder reviews are limited to 20 files per command to maintain context quality. Large diffs exceeding model context windows may require chunking.
Folder reviews are limited to 20 files at a time. Requires Git for branch and staged diff functionality.
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 any payment.
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
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 2 days ago
- Free with an Agensi account
Requires
Permissions
Allowed Hosts