xby-review-code
Run the following command in DeepSeek Harness:
dsh plugin install xby-skill/xby-review-code
Paste the following prompt into your AI chat to install this plugin:
Visit https://github.com/xby-skill/xby-review-code to obtain the source code, then run dsh plugin install xby-skill/xby-review-code in the DeepSeek Harness terminal to complete the installation.
About this plugin
AI-assisted development is everywhere, yet code quality assessment still lacks a structured, quantifiable approach. xby-review-code fills that gap: built on the Model Context Protocol, it does not call an LLM directly. Instead, it crafts high-quality, multi-dimensional review prompts that agents can reuse and compare across iterations.
The plugin supports three review granularities — full codebase, Git diff, and single file — each paired with scoring dimensions. Combined with the parse_review_score tool, it can extract a quantified result from review text, closing the loop from prompt construction to score extraction. It also offers a set_xby_apikey tool for persistent API key management across restarts.
If you are wiring code-quality gates into an agent workflow, embedding code review into an automation pipeline, or need multi-dimensional scores to compare alternative implementations, this plugin is ready to use. Simply obtain an API key from the service provider and invoke it inside DeepSeek Harness.
Use Cases
- Multi-dimensional code quality assessment within agent workflows
- Building Git diff review prompts and extracting quantified scores in automation pipelines
- Obtaining structured single-file scores when comparing alternative implementations
Best For
- Developers embedding code-quality gates into agent workflows
- Tech leads needing multi-dimensional scores to compare architectural options
- Platform engineers building automated code review pipelines
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.