mcp_analysis 0.1.0
mcp_analysis: ^0.1.0 copied to clipboard
Analysis engine for the MCP ecosystem providing pipeline-based data analysis with pluggable functions, transforms, and multi-source data integration.
MCP Analysis #
Analysis engine for the MCP ecosystem. Pipeline-based data analysis with pluggable functions, transforms, and multi-source data integration. Implements the AnalysisPort / AnalysisFunctionPort / AnalysisDataSourcePort Contract Layer defined in mcp_bundle.
Components #
- Spec — spec validator, parameter resolver, spec manager.
- Artifact — artifact builder, store, provenance tracker.
- Execution — execution engine, batch / ad-hoc / stream executors, job manager, retry policy, step logger.
- DataSource — multi-source data integration with pluggable adapters.
- Functions and Transforms — pluggable analysis functions and data transforms.
- Alerts — alert evaluation against analysis results.
- MCP integration — exposes analysis capabilities through MCP tools.
- Standard port adapter —
analysis_port_adapter.dartimplementing themcp_bundleanalysis Contract Layer.
Quick Start #
import 'package:mcp_analysis/mcp_analysis.dart';
final engine = ExecutionEngine(
jobManager: JobManager(),
artifactStore: InMemoryArtifactStore(),
);
final result = await engine.executeAdhoc(spec, parameters);
Support #
License #
MIT — see LICENSE.