CodeForgeX
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
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 |
|---|---|
| list_filesA | List files and directories in the repository. Args: directory: Relative subfolder inside the repository (empty string for root). recursive: Whether to list recursively through child directories (default: True). max_depth: Maximum directory recursion depth (default: 10). Returns: JSON string containing total_entries and a structured list of files with sizes. |
| read_fileA | Read text content from a file inside the repository with line windowing. Args: path: Relative path to the file inside the repository. start_line: 1-indexed starting line number (default: 1). end_line: 1-indexed ending line number (default: None, reads up to 1000 lines). Returns: JSON string containing path, line numbers, content, and truncation status. |
| search_codeA | Search for literal text or regular expressions across repository files. Args: query: The substring or regex to search for. path: Optional relative directory or file to constrain search scope. is_regex: Whether query is a regular expression (default: False). case_sensitive: Whether match should be case-sensitive (default: False). max_results: Maximum matching lines to return (default: 50). Returns: JSON string with matched lines, line numbers, and file paths. |
| run_testsA | Execute automated pytest tests within the repository sandbox. Args: test_target: Optional relative test file or test node (e.g. 'tests/test_math.py::test_add'). timeout_seconds: Maximum time allowed before terminating process (default: 30s). Returns: JSON string with exit code, passed/failed counts, duration, and output summary. |
| get_test_outputA | Retrieve test execution output and logs from a previous test run. Args: run_id: Specific test run identifier (empty string defaults to most recent run). full: If True, returns full untruncated stdout and stderr (default: False). Returns: JSON string containing logs, exit code, duration, and pass/fail counts. |
| apply_patchA | Atomically apply a unified diff patch to repository files. Args: patch: The unified diff content (e.g., standard 'diff --git' or '--- / +++' format). file_path: Optional relative target path if targeting a single file. Returns: JSON string with success status, list of changed files, and status message. |
| get_git_diffA | Retrieve the deterministic Git diff of changes made in the repository. Args: path: Optional relative file or directory path to scope the diff. cached: If True, inspects staged changes (git diff --cached) (default: False). Returns: JSON string containing diff string, has_changes boolean, and changed_files list. |
| get_repository_statusA | Inspect the current Git repository status (modified, staged, untracked files). Returns: JSON string with branch, clean flag, and lists of modified/staged/untracked files. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| explore_repository_prompt | Prompt template guiding an agent to explore a repository systematically. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| get_repo_overview | Resource providing high-level repository metadata and bounded root location. |
TDQS
Scored across 8 tools
Each tool targets a distinct concern: file listing, status inspection, content reading, search, test execution, test output retrieval, patching, and diff viewing. The only related pair is run_tests/get_test_output, but their roles are clearly sequential rather than interchangeable.
Most tools follow a clear verb_noun pattern like list_files, read_file, search_code, run_tests, and apply_patch. The get_* variants are slightly different in style but still predictable and consistent with common Git-related naming.
Eight tools is a well-scoped size for a repository coding assistant. Each tool covers a necessary operation without redundancy or bloat, and the set is small enough for an agent to navigate easily.
The toolset covers the core inspect-modify-test loop well: reading, searching, patching, diffing, and running tests. Missing commit/branch management tools are a minor gap for full repository lifecycle coverage, but agents can still complete most coding workflows.