Yellhorn MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| REPO_PATH | No | Path to your repository | current directory |
| GEMINI_API_KEY | Yes | Your Gemini API key | |
| YELLHORN_MCP_MODEL | No | Gemini model to use | gemini-2.5-pro-exp-03-25 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| create_workplanA | Creates a GitHub issue with a detailed implementation plan. This tool will:
The AI will analyze your entire codebase (respecting .gitignore) to create a detailed plan with:
Codebase reasoning modes:
Returns the created issue URL and number immediately. |
| get_workplanB | Retrieves the workplan content (GitHub issue body) for a specified issue number. |
| revise_workplanA | Updates an existing workplan based on revision instructions. This tool will:
The AI will use the same codebase analysis mode and model as the original workplan. Returns the issue URL and number immediately. |
| curate_contextA | Analyzes the codebase and creates a .yellhorncontext file listing directories to be included in AI context. This tool helps optimize AI context by:
The .yellhorncontext file acts as a whitelist - only files matching the patterns will be included. This significantly reduces token usage and improves AI focus on relevant code. Example .yellhorncontext: src/api/ src/models/ tests/api/ *.config.js |
| judge_workplanA | Triggers an asynchronous code judgement comparing two git refs against a workplan. This tool will:
The judgement will evaluate:
Supports comparing:
Returns the sub-issue URL immediately. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool has a distinct primary purpose: create_workplan initiates planning, curate_context manages codebase context, get_workplan retrieves plans, judge_workplan evaluates implementations, and revise_workplan updates plans. There is minor potential overlap between create_workplan and revise_workplan (both generate/update workplans), but their distinct triggers and descriptions help differentiate them effectively.
All tools follow a consistent verb_noun naming pattern (e.g., create_workplan, judge_workplan, revise_workplan). The verbs are clear and descriptive, and the snake_case style is uniformly applied across all five tools, making the set predictable and easy to navigate.
With 5 tools, the server is well-scoped for its purpose of managing AI-driven workplan creation and evaluation in GitHub contexts. Each tool serves a specific, necessary function in the workflow, from planning to context curation to judgement, without redundancy or bloat, fitting a typical range for focused MCP servers.
The tool set covers the core lifecycle of workplan management: creation (create_workplan), retrieval (get_workplan), revision (revise_workplan), and evaluation (judge_workplan), with context optimization (curate_context) as a supporting function. A minor gap exists in direct deletion or archiving of workplans, but agents can likely handle this through GitHub's native tools, and the coverage supports end-to-end workflows effectively.