Loading...
Analyzes codebase structure, ranks file importance, tracks dependencies, and provides summaries for enhanced code understanding by AI assistants.
Boost this tool
Subscribe to listing upgrades or segmented pushes.
Analyzes codebase structure, ranks file importance, tracks dependencies, and provides summaries for enhanced code understanding by AI assistants.
The FileScopeMCP server is relatively safe for read-only operations and code analysis. Risks are elevated when enabling file watching or when AI is used to generate summaries, as this could expose sensitive information. Ensure the base directory is properly scoped and file watching is configured with appropriate exclusions.
Performance depends on the size of the codebase and the complexity of the dependencies. File watching can introduce overhead, especially on large projects. Caching is used to improve performance for repeated operations.
The FileScopeMCP server is free to use. However, generating diagrams and analyzing large codebases may require significant computational resources.
{
"mcpServers": {
"FileScopeMCP": {
"command": "node",
"args": ["<build script sets this>/mcp-server.js","--base-dir=C:/Users/admica/my/project/base"],
"transport": "stdio",
"disabled": false,
"alwaysAllow": []
}
}
}list_saved_treesLists all saved file tree configurations.
Read-only operation, no modification of data.
create_file_treeCreates a new file tree configuration for a specific directory.
Creates a new configuration file, but does not access sensitive data.
select_file_treeSelects an existing file tree to work with.
Read-only operation, selects a configuration file.
delete_file_treeDeletes a file tree configuration.
Deletes a configuration file, potentially losing analysis data.
list_filesLists all files in the project with their importance rankings.
Read-only operation, no modification of data.
get_file_importanceGets detailed information about a specific file, including dependencies and dependents.
Read-only operation, no modification of data.
find_important_filesFinds the most important files in the project based on configurable criteria.
Read-only operation, no modification of data.
read_file_contentReads the content of a specific file.
Read-only operation, accesses file content.
recalculate_importanceRecalculates importance values for all files based on dependencies.
Read-only operation, recalculates values based on existing data.
get_file_summaryGets the stored summary of a specific file.
Read-only operation, retrieves stored summary.
set_file_summarySets or updates the summary of a specific file.
Writes a summary to the configuration file.
toggle_file_watchingToggles file watching on/off.
Modifies the file watching configuration.
get_file_watching_statusGets the current status of file watching.
Read-only operation, retrieves file watching status.
update_file_watching_configUpdates file watching configuration.
Modifies the file watching configuration.
generate_diagramCreates Mermaid diagrams with customizable options.
Generates diagram files based on project structure.
None
local
The FileScopeMCP server is relatively safe for read-only operations and code analysis. Risks are elevated when enabling file watching or when AI is used to generate summaries, as this could expose sensitive information. Ensure the base directory is properly scoped and file watching is configured with appropriate exclusions.
Autonomy is limited by the file system permissions of the user running the server. Exercise caution when granting write access to the configuration files.
Production Tip
Monitor file system usage and resource consumption when file watching is enabled to prevent performance issues.
File importance is calculated based on a weighted formula that considers the number of incoming and outgoing dependencies, file type, location, and naming conventions.
The server supports Python, JavaScript, TypeScript, C/C++, Rust, Lua, Zig, C#, and Java.
Exclusions are handled automatically by ignoring hidden virtual environments and other common unwanted directories. Custom exclusions can be implemented by modifying the code.
Yes, you can customize the diagram style, depth, minimum importance threshold, layout, and output format.
You can enable file watching using the `toggle_file_watching` tool. You can also configure file watching options using the `update_file_watching_config` tool.
File tree data is stored in JSON files in the file system.
No, the server is single-threaded.