dsh-pr-description
Run the following command in DeepSeek Harness:
dsh plugin install 988hj7tczd-oss/dsh-pr-description
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install 988hj7tczd-oss/dsh-pr-description in the DeepSeek Harness terminal to install; browse the source at https://github.com/988hj7tczd-oss/dsh-pr-description .
About this plugin
Hand-writing PR descriptions is repetitive and often misses testing or risk notes, while inconsistent formatting across the team hurts review quality. dsh-pr-description analyzes the current branch diff, classifies changes into feat/fix/refactor/docs/test/chore via deterministic rules, extracts new symbols, and assembles a Conventional Commits title plus five sections (Motivation, Approach, Impact, Test Suggestions, Risks & Open Questions) along with a four-item self-review checklist. Output is written as Markdown into the workspace and rendered inline. The pipeline is fully offline with no LLM calls. Optional confirm prompts for user approval and openPr invokes the gh CLI to create a Pull Request directly.
Best for DSH developers who submit PRs frequently and value consistent team formatting, controlled command execution, and testability. All git and gh invocations use fixed argv arrays rather than shell strings, and every parsing, classification, and templating step is a zero-dependency pure function. Subprocess trees are cleaned up automatically on plugin unload.
Use Cases
- Generating a Conventional Commits title and five-section PR body before submission
- Enforcing consistent PR formatting with a self-review checklist across the team
- Producing PR description text fully offline without any LLM dependency
Best For
- DSH workflow developers who submit PRs frequently
- Engineers who value commit conventions, controlled execution, and testability
- Smaller tech teams looking to cut down on repetitive PR documentation work
Related Plugins
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.
An armor-breaking plugin for DeepSeek: rewrites instruction priority, expands refusal-blocked phrases, persists across sessions, and shows a green active indicator.