#1414th of 15 in Code review and testing
ai-codereview-gitlab
LLM reviewer for GitLab merge requests and pushes, with chat notifications
- Stars
- 1.9k
- License
- Apache-2.0
- Last commit
- Oct 2026
- Last release
- Jun 2026
- Language
- Python
1,551 on 31 Mar 20261,863 on 20 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Receives GitLab webhooks for Push and Merge Request events, sends the diff to a configured LLM, and posts the review back as a Note on the MR or commit. It also pushes results to DingTalk, WeCom or Feishu, and includes a Streamlit dashboard and daily reports built from GitLab, GitHub and Gitea commits. An optional agentic mode lets the model read files and run read-only shell commands in a local clone.
Who it is for: Teams on GitLab wanting self-hosted LLM review of merge requests and pushes
Strengths
- Supports DeepSeek, ZhipuAI, OpenAI, Anthropic, Qwen and Ollama via one LLM_PROVIDER setting
- Agentic mode falls back to diff_only on clone, fetch, LLM or tool failure
- Agent shell is sandboxed: read-only allowlist, path checks, 30s timeout
- Includes a dashboard on port 5002 with project and developer statistics
Weaknesses
- README and docs are mainly in Chinese
- Webhook flow is documented for GitLab only; GitHub and Gitea appear only in daily reports
- Agentic mode needs 50GB+ disk and uses 3-10x the tokens of diff_only
- Some features sit in a separate Pro edition
What it needs
- no GPU
- Docker + Compose
- Needs GitLab access token, LLM provider API key or Ollama, DingTalk, WeCom or Feishu webhook (optional)
- Models: DeepSeek, ZhipuAI, OpenAI, Anthropic, Qwen
- port 5001
Also in Code review and testing
See all 15| Rank | Project | Score |
|---|---|---|
| 1 | open-code-reviewCLI that reviews Git diffs with an LLM agent and line-level comments | 80 out of 100 |
| 2 | react-doctorCLI that scans React codebases for state, performance, security and accessibility issues | 74 out of 100 |
| 3 | hunkTerminal diff viewer for reviewing code changes, with inline agent notes | 74 out of 100 |
| #4 | no-mistakesLocal git proxy that runs AI review, tests and lint before opening a PR | 73 out of 100 |
| #5 | pr-agentLLM pull request reviewer that comments via /review, /describe and /improve | 71 out of 100 |