Code Generator MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CODE_GEN_MODEL | No | Model name (alternative: OPENAI_MODEL). Default: gpt-4o | gpt-4o |
| CODE_GEN_API_KEY | No | API key (alternative: OPENAI_API_KEY). Optional for local endpoints. | |
| CODE_GEN_API_URL | No | OpenAI-compatible endpoint URL (alternative: OPENAI_BASE_URL). Default: https://api.openai.com/v1 | https://api.openai.com/v1 |
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
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| generate_standard_functionC | Generate code for a standard function spec (Template 1) using coder expert model. If IMPORTANT: All parameters (such as 'task', 'description', 'constraints', 'edge_cases', etc.) MUST be provided in English. If the user's prompt or request is in Vietnamese or another language, the calling agent must automatically translate the text of these parameter values into English before invoking this tool. |
| generate_codebase_contextC | Generate code for a function with codebase context (Template 2) using coder expert model. If IMPORTANT: All parameters (such as 'task', 'description', 'constraints', etc.) MUST be provided in English. If the user's prompt or request is in Vietnamese or another language, the calling agent must automatically translate the text of these parameter values into English before invoking this tool. |
| generate_bugfix_refactorC | Generate fixed or refactored code (Template 3) using coder expert model. If IMPORTANT: All parameters (such as 'task', 'problem', 'expected_behavior', 'constraints', etc.) MUST be provided in English. If the user's prompt or request is in Vietnamese or another language, the calling agent must automatically translate the text of these parameter values into English before invoking this tool. |
| generate_multi_function_moduleB | Generate code for a multi-function module (Template 4) using coder expert model. If IMPORTANT: All parameters (such as 'task', 'module_purpose', 'context', 'functions' list with its descriptions/constraints, etc.) MUST be provided in English. If the user's prompt or request is in Vietnamese or another language, the calling agent must automatically translate the text of these parameter values into English before invoking this tool. |
| run_project_testsA | Automatically detects and runs the project's test suite, returning the command's stdout and stderr. Can be used by the AI to verify correct behavior of generated code. If 'custom_command' is provided, it runs that command line instead of auto-detecting. |
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
The four generate_* tools map to distinct templates/scenarios (standard function, codebase-aware function, bugfix/refactor, multi-function module), so an agent can usually pick correctly. However, generate_standard_function and generate_codebase_context both produce a single function and differ only by whether codebase context is used, which could cause occasional misselection.
All names use snake_case with a consistent verb_noun pattern (generate_* plus run_project_tests). The generate_* family is predictable and self-describing.
Five tools is well-scoped for a code-generation server: four generation templates plus a test runner, each earning its place. No redundant or filler tools.
The surface covers code generation across common scenarios and test execution/verification, with test suites co-generated. Minor gaps exist around applying/writing generated code back to disk or editing existing files, but core workflows are covered.