Fitbit Health Skill

• Outcome: Pass
• Latest audit: May 27, 2026, 9:19 PM
• Version:
  0.1.1

Overview

This is a coherent Fitbit CLI skill that accesses sensitive health data with disclosed OAuth setup and local token storage, so users should install it only if they are comfortable sharing Fitbit data with their agent.

Before installing, review the Fitbit consent screen carefully because the skill requests access to activity, heart-rate, sleep, weight, and profile scopes. Use it only in conversations where Fitbit health details may safely appear, keep the local token file private, and run `fitbit logout` or revoke the app in Fitbit when you no longer need it.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 91% confidence
• Finding: The skill invokes a CLI that performs network access to Fitbit APIs and shell execution, yet it declares no explicit permissions or user-facing consent boundaries. In an agent environment, this can lead to unexpected external requests and handling of sensitive health/account data without clear authorization semantics, increasing the chance of overreach or accidental data exposure.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 95% confidence
• Finding: The description frames the skill as a simple Fitbit data query tool, but the documented behavior includes OAuth setup/login, token lifecycle handling, and access to broader profile/account information. That mismatch is dangerous because users or orchestrators may approve the skill expecting limited health summaries while it can authenticate accounts and retrieve additional sensitive personal data.

Missing User Warnings

Medium

• Confidence: 89% confidence
• Finding: The README explicitly promotes connecting an AI agent to Fitbit health data, which is sensitive personal information, but it does not warn users about privacy risks, data minimization, retention, or who may receive that data once surfaced through the agent. In this context, omission of privacy guidance can lead users to expose health metrics to the agent ecosystem or downstream logs/integrations without informed consent, increasing the chance of unintended disclosure.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: This skill accesses highly sensitive health information, but the description does not warn users that using it may expose personal activity, sleep, heart-rate, weight, and profile data. Lack of an explicit warning increases the risk of uninformed consent, accidental disclosure in chat transcripts, and unsafe use in shared or logged environments.

VirusTotal

65/65 vendors flagged this skill as clean.