Art-forward terminal AI agent with OpenAI-compatible tools and an aurora TUI. Supports MCP (Model Context Protocol) servers for tool integration.
VORTEX exposes a single tool 'mcp_tool' with a completely empty input schema (no properties, no required fields). The tool description is generic and does not explain what parameters it accepts, how to invoke it, or what it returns. The schema property method in mcp_tool.py extracts properties from _tool_info.input_schema, but the tool itself has no documented input contract. This violates the core requirement that every tool must have a non-empty description and a complete input schema with typed parameters. The tool is a wrapper around dynamically loaded MCP tools, but the wrapper itself is undocumented and unusable by an LLM without runtime introspection.
Executes tools from connected MCP (Model Context Protocol) servers. Dynamically wraps MCP tool definitions and executes them with provided parameters.
Tool input schema is completely empty: {"type":"object","properties":{},"required":[]}. LLMs cannot determine what parameters to pass or how to invoke the tool.
Tool description is generic and does not explain WHAT the tool does, WHEN to use it, or WHAT it returns. Current description: 'Executes tools from connected MCP (Model Context Protocol) servers. Dynamically wraps MCP tool definitions and executes them with provided parameters.', lacks actionable guidance for LLM selection.
Tool name 'mcp_tool' is generic and does not start with an action verb. It does not convey what action the tool performs. Rename to reflect the actual capability (e.g., 'execute_mcp_tool' or 'call_remote_tool').
No parameter descriptions provided. The tool accepts arbitrary parameters via invocation.params, but there is no documentation of what parameters are valid, their types, constraints, or dependencies.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 22 | 2026-07-28+ | v2 |
No output schema documented. The tool returns ToolResult with either success or error, but the structure of the output, expected fields, and data types are not documented for the LLM.