skills.rest
VMAFx
VMAFxOFFICIAL
Explore all skills by this author →
WebsiteGitHub

dev-llm-review

Run a local LLM code review on a single file with vmaf-dev-llm.

3|1|Updated May 28, 2026
Search Tags:#code review#read-only#ollama#local llm#pre-merge review#vmaf-dev-llm
One-click install
npx skills add https://github.com/VMAFx/vmafx --skill dev-llm-review
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: dev-llm-review
Source: https://github.com/VMAFx/vmafx/tree/main/.claude/skills/dev-llm-review
Command: npx skills add https://github.com/VMAFx/vmafx --skill dev-llm-review

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps developers run a fast, local code review on a specific file before opening a pull request, reducing the need for ad hoc manual review passes.

Core Features & Use Cases

  • Local Review Workflow: Checks that the review tool is installed and that Ollama is reachable before running the review.
  • File-Specific Analysis: Reviews exactly one target file and can optionally switch to a named model for different review styles.
  • Safety Guardrails: Keeps the review read-only, prints the model output verbatim, and refuses vendored upstream paths under subprojects.
  • Use Case: Use it when you want a local self-review of a changed source file without sending code to a cloud service.

Quick Start

Use the dev-llm-review skill to review src/main.py with the qwen2.5-coder:7b model.

Frequently Asked Questions about dev-llm-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a local code review on a single file before opening a pull request?▼

You can run a local code review on a single file before a pull request by using a local LLM with a reachable Ollama backend. The review executes read-only, prints the model output verbatim, and refuses vendored subproject paths.

Do I need a cloud service for pre-merge LLM code review?▼

No, you do not need a cloud service for pre-merge LLM code review if you use a local Ollama backend. The tool checks that the review CLI is installed and that Ollama is reachable, then runs the analysis locally.

How does local code review handle vendored upstream paths?▼

Local code review handles vendored upstream paths by refusing to review files under vendored subproject paths. This safety guardrail prevents accidental analysis of third-party upstream code.

Can I use a different model to review source files locally for different review styles?▼

Yes, you can review source files locally with a different model for different review styles by providing an optional model override. This allows you to switch the named model used for the file-specific analysis.

What are the limitations of running a local LLM code review?▼

Limitations of running a local LLM code review include being restricted to read-only execution and analyzing exactly one target file per run. It also requires a locally installed CLI and refuses vendored subproject paths.

Why does my local code review require a reachable Ollama backend?▼

Your local code review requires a reachable Ollama backend because the tool relies on local LLMs to generate the analysis. The workflow checks that Ollama is reachable before running the file-specific review.

Related Skills

Explore More Software Engineering Skills→
MacPhobosMacPhobos

Local LLM Ops (Ollama)

Set up Ollama and manage local LLMs on Apple Silicon.

Community
Intermediate
khokskhoks
2

change_management

Route tracked file modifications through pull requests for owner review.

Community
Advanced
dennisonbertramdennisonbertram
1

ollama-ops

Orchestrate local LLM inference and model management with Ollama.

Community
Intermediate
Exia-thdExia-thd

code-reviewer

Analyze source code and produce severity-ranked review reports with metrics and auto-fix suggestions.

Community
Advanced