Comprehensive MCP server providing systematic tools and references for Python-to-TypeScript porting with real-world examples, optimized for Python 3.9+ codebases. Includes real execution, differential testing, AST analysis, orchestrated workflows, and hybrid architecture support.
This MCP server exhibits significant definition quality gaps across nearly all 10 tools. While tool names follow verb-noun patterns (e.g., 'porting-strategy', 'type-analysis'), the actual tool definitions are problematic: descriptions are present but generic and lack LLM-optimization guidance; input schemas are declared but lack detailed parameter descriptions and proper type constraints; output schemas are entirely undocumented; error handling is absent. The server relies on inferred tool definitions from src/resources/server-capabilities.ts, where tools are listed as strings in arrays rather than explicitly registered with full schemas. No evidence of input validation, parameter constraints, or error recovery guidance. The tools operate at a domain level (Python-to-TypeScript porting) but their definitions do not meet production quality standards for agent orchestration.
Add a step to an active porting notebook with analysis and implementation details
Create an ephemeral Srcbook journal for structured porting work and documentation
Compare Python and TypeScript implementations to verify correctness of ported code through differential testing
Execute TypeScript code in a sandboxed environment and return results with memory and performance metrics
Find TypeScript/JavaScript equivalents for Python libraries with installation commands and migration guidance.
Convert Python language patterns and idioms to TypeScript equivalents with examples and caveats. Emphasizes Python 3.9+ modern patterns including union operators, dict merge operators, and built-in generics.
Tool definitions are inferred from string arrays rather than explicitly registered with full schemas. Only tool names and descriptions visible; no actual JSON Schema objects or formal registration code shown.
Parameter descriptions are present in schema declarations but are extremely generic ('Information about the Python project', 'Constraints and requirements'). LLMs cannot infer specific formats, ranges, enums, or valid values from these. No parameter constraints, regex patterns, or length limits documented.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 45 | <=2025-11-25 | v2 |
| 2026-03-09 | F | 38 | - | v1 |
Generate systematic porting strategies for Python projects
Analyze Python code using tree-sitter to extract AST structure, complexity metrics, and migration recommendations
Analyze Python type hints and patterns for TypeScript conversion with Python 3.9+ optimization
Generate validation and testing strategies for ported TypeScript code
Output schemas are completely undocumented. None of the 10 tools include return type definitions or field descriptions. Agents cannot plan downstream tool calls or parse results reliably.
No error handling, recovery guidance, or actionable error messages defined. Tools marked as READ_ONLY or WRITE risk but no error taxonomy, retryability classification, or user-fixable vs fatal distinctions provided.
Tools with WRITE risk ('create-porting-notebook', 'add-porting-step', 'execute-notebook-cell', 'differential-test') have no dry-run, confirmation, or idempotency guarantees documented. Agents may execute destructive operations without safeguards.
No tool composition hints. It is unclear how agents should chain tools together. E.g., should 'porting-strategy' output be fed into 'type-analysis'? Does 'create-porting-notebook' return a notebookId that 'add-porting-step' requires? Broken dependency chains force agents to guess.
Parameter enum 'analysisLevel' in 'python-ast-analyze' is declared with enum constraint, but no description of what 'basic' vs 'comprehensive' means. LLMs cannot reason about which level to select without guidance.
Tool descriptions do not state WHEN to use each tool vs similar alternatives. E.g., 'type-analysis' vs 'python-ast-analyze', both analyze Python code. The distinction is unclear. Agents will conflate them.