Loading ToolBench

Reading the current corpus…

Portfolio MCP server · F · ToolBench

Portfolio

A personal portfolio website built with React, TypeScript, and Vite. Features a chatbot powered by Gemini API, project showcase, certifications, CV modal, and contact forms.

2026-07-28+1 tools

14/100

Definition Quality12

Protocol Readiness0

Supportability25

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence

Current-spec patterns detected

Summary

This is not an MCP server. The repository is a React portfolio website (Vite + TypeScript) with a chatbot UI component. The 'sendMessage' tool is a frontend function in src/components/chatBot/api/geminiCall.ts that calls the Gemini API directly from the browser. There is no MCP server implementation, no tool registration via MCP protocol, no schema definitions, and no transport layer. The tool lacks input validation, error handling, and structured output. This submission appears to be a misclassification, it is a web application, not an MCP server.

Tools (1)

sendMessageread onlyauth15/100

Sends a message to the Gemini API and returns the model's response for the chatbot

Issues (5)

CRITICALtool ↗

Not an MCP server. This is a React web application with a frontend chatbot component, not a Model Context Protocol server.

sendMessage

CRITICALconstrained-input ↗

No input schema defined. The tool accepts 'messages' array but provides no JSON Schema with type definitions, constraints, or field descriptions.

sendMessage

CRITICALtool ↗

No output schema documented. Callers cannot know what fields the Gemini response contains or how to parse it.

sendMessage

HIGHrecovery-guide ↗

No error handling or recovery guidance. If the Gemini API fails, there is no structured error response or guidance for the caller.

sendMessage

HIGHsecret-injection ↗

API key exposure risk. The Gemini API key is likely embedded in frontend code or environment variables accessible to the browser, violating secret-injection pattern.

Recommendations

This repository must be restructured as an actual MCP server. Use the official MCP SDK (Node.js or Python) to implement a proper server with stdio or HTTP transport.
Define a formal JSON Schema for the 'messages' parameter with type 'array', items with 'role' enum (user|model) and 'text' string, and add descriptions for each field.
Document the output schema returned by Gemini, including response fields, types, and any pagination or metadata.
Implement structured error handling that returns error codes, messages, and recovery hints (e.g., 'API rate limit exceeded. Retry after 60 seconds.').
Move API key management to server-side environment variables or a secrets vault. Never expose credentials in frontend code or tool parameters.

Score history

Overall score trend

First recorded score · v2 rubric

14/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	14	2026-07-28+	v2