cribl-mcp

# Changelog All notable changes to this project will be documented in this file. ## [Unreleased] ## [0.1.12] - 2025-05-07 ### Added - `cribl_versionControl`: Detects if version control is enabled. - `cribl_commitPipeline`: Commits staged pipeline config changes. - `cribl_deployPipeline`: Deploys the latest committed configuration to a worker group. ### Changed - Restored placeholder description for `groupName` in all Cribl tool schemas by reverting to `z.string().optional()`, removing `nullable()` and default values so the inspector UI shows the description as placeholder. - Use `z.preprocess` to map `null` to `undefined` for optional `groupName` argument, ensuring Inspector's blank input triggers fallback while showing placeholder text. - Use `z.preprocess` to map `null`, `undefined`, or empty string to default `'stream'` for `productType` argument in `cribl_listWorkerGroups`, ensuring Inspector's blank input uses the default. ## [0.1.11] - 2025-04-19 ### Fixed - Re-aligned versioning with npm registry to correct earlier tagging inconsistency - Cleaned up publishing flow to ensure only scoped version (`@pebbletek/cribl-mcp`) is used going forward ### Added - GitHub release notes - Deprecation notice for unscoped `cribl-mcp` package ## [0.1.10] - 2025-04-19 ### Changed - Reverted .env loading logic to support MCP_DOTENV_PATH override and fallback to default search - Added descriptive summaries for all Cribl MCP tools in `src/server.ts` ## [0.1.9] - 2025-04-18 ### Added - Documented support for `MCP_DOTENV_PATH` environment variable in `src/config.ts` to specify a custom `.env` file location. See `docs/configuration.md` and `docs/usage.md` for examples. - Documented running the server via `node dist/server.js` in `docs/usage.md`. - Added `cribl_getSystemMetrics` tool and optional parameters (`filter`, `metrics`, `earliest`, `latest`, `buckets`) to allow filtering and scoping of metrics queries. - Updated `cribl_getSystemMetrics` tool description to strongly recommend using parameters to limit response size and clarify default of `buckets=1`. ### Changed - Removed `cribl_getPipelineStats` from the "Future Enhancements" section in `docs/overview.md` as `cribl_getSystemMetrics` covers this functionality. - Renamed `filter` parameter to `filterExpr` and added `wp` parameter to `cribl_getSystemMetrics` tool and documentation for consistency with Cribl API. - Updated `cribl_getSystemMetrics` tool and documentation to use official Cribl API parameter names: `buckets` is now `numBuckets`, and `metrics` is now `metricNameFilter`. ### Fixed - Adjusted error messages in `src/config.ts` related to missing environment variables to mention `MCP_DOTENV_PATH`. ## [0.1.5] - 2024-06-21 ### Added - Added shebang (`#!/usr/bin/env node`) to `src/server.ts` to ensure it can be executed directly. ### Changed - Modified `build` script in `package.json` (`tsc && chmod +x dist/server.js`) to make the compiled `dist/server.js` executable after each build. - Removed MacOS related files (.DS_Store) and updated .gitignore. - Updated documentation (`README.md`, `docs/usage.md`) to use the correct `npx` organization name (`@pebbletek/cribl-mcp` instead of `@pebble/cribl-mcp`) and remove outdated references to `fastmcp`. - Improved cross-linking between documentation files (`README.md`, `docs/overview.md`, `docs/usage.md`, `docs/configuration.md`). - Updated `README.md` and `docs/usage.md` with clearer instructions on running via `npx` and handling environment variables directly on the command line. - Updated documentation to reflect that `mcp_Cribl_cribl_restartWorkerGroup` no longer requires a dummy parameter and takes no arguments. ## [0.1.4] - 2025-04-16 ### Changed - Updated `package.json` after it became corrupted. ## [0.1.2] - 2025-04-15 ### Added - Dynamic authentication supporting Cribl.Cloud (Client ID/Secret) and Local (Username/Password) via `.env` settings (`CRIBL_AUTH_TYPE`). - Automatic token acquisition and refresh using Axios interceptor. - `cribl_listWorkerGroups` tool to list groups, with optional `productType` filtering (stream, edge, search, all). - `cribl_getPipelineConfig` tool to fetch configuration for a specific pipeline. - Automatic Worker Group selection: If `groupName` is omitted for relevant tools, attempts to use the sole 'stream' group if only one exists. - More robust error handling in API client (`