submit: kestarsheng-code-review-agent

[kestarsheng]

Code Review Agent

Capability

Dual-engine AI code review service: a deterministic rule engine (26 rules across Python, JavaScript, Java, Go, Rust) plus LLM semantic analysis with cross-validation. Returns a structured quality report with 5-dimension scores (correctness, security, performance, maintainability, best practice), issues with source attribution (rule/llm/confirmed), and directly applicable fix code.

Live API

• Base URL: https://code-review-agent-ashy-six.vercel.app/v1
• Health Check: https://code-review-agent-ashy-six.vercel.app/health
• Verification: https://code-review-agent-ashy-six.vercel.app/.well-known/xagent-verification.json
• Authentication: none

Tech Stack

• FastAPI + OpenAI-compatible protocol (DeepSeek)
• FastMCP Server - 7 MCP tools, streamable HTTP transport at /mcp
• Deployed on Vercel (serverless, push-to-deploy)

Key Features

• POST /v1/review - dual-engine review of a code snippet (+ optional language/context)
• POST /v1/review_diff - review a unified diff with change metadata
• POST /v1/review_files - multi-file batch review (per-file scan + cross-file analysis)
• POST /v1/suggest_fix - full corrected code + change notes
• GET /v1/rules - list all 26 rule-engine rules
• 5-dimension scoring (0-100 each) + weighted composite score (A-D grade)
• Each issue carries fix_code (rule engine auto-generates for 8 key rule types)
• MCP endpoint at /mcp with 7 tools (review_code, review_diff, review_files_tool, suggest_fix, explain_issue, list_rules, get_rule_detail)
• CLI (cli.py) for git-diff review: python cli.py --commit HEAD~1
• Safe error handling: 422 (empty body), 413 (oversized), 502 (LLM failure)

Source

• Repository: https://github.com/kestarsheng/code-review-agent
• Review commit: a4b5d3d644a0a0bcd0ca7db780671f47d77fedc8
• Full source included in source/

Builder

kestarsheng - 2410251355@henu.edu.cn

[github-actions]

Submission received — #50

Thank you for submitting to the current X-Agent MCP Hackathon. We have received your pull request.

This is a receipt only. It does not confirm eligibility, completeness, technical approval, a judging result, or an award. Passing automated checks or merging code for archival purposes does not, by itself, mean the entry has passed review or won a prize.

Please include the complete source for the submitted capability in this repository under submissions/mcp-hackathon/<slug>/source/, with dependency files, configuration examples, and reproducible setup/build instructions. An external repository link alone is not sufficient. The official copy is intended to remain available if the original repository later becomes unavailable. Do not include secrets, credentials, or private user data; clearly disclose external and private-service dependencies.

The automated checks report technical checks separately. A workflow failure is not a judging decision. Keep updates in this PR while submissions remain open, and keep the documented source version and deployment evidence consistent.

See the submission instructions. Final review results and awards will be announced separately through the official event channels.