Back to skill

Security audit

garmin-health-analysis

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this Garmin health analysis skill suspicious because its example exposes a password in a command line, its trigger may retrieve health data for an ambiguous question, and a profile command includes account identity information; the scan was partial.

Review the credential example, broad trigger, and profile data behavior before installing. The scanner recommends not installing, and its analysis was partial because three local references could not be resolved; the clean AIG result does not remove these concerns or establish that the skill is safe.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (6)

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · references/mcp_setup.md (reported line 28)May include surrounding context.

md
# Install and setup
npm install
pip3 install garminconnect fitparse gpxpy
cp .env.example .env
# Edit .env with your credentials

# Authenticate

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · references/mcp_setup.md (reported line 29)May include surrounding context.

md
npm install
pip3 install garminconnect fitparse gpxpy
cp .env.example .env
# Edit .env with your credentials

# Authenticate
npm run auth

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, file_read, file_write, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The command-line authentication example exposes the password without warning about shell history or process-list visibility.

Medium
Category
Output Handling
Confidence
99% confidence
Finding

Following this example places a real Garmin password in command-line arguments, potentially exposing it through shell history, process inspection, or captured execution logs. Compromise of this credential can expose sensitive health and activity information, making the otherwise legitimate authentication workflow security-relevant.

Content

No source excerpt is available for this finding.

The trigger “how am I doing?” can initiate Garmin health-data retrieval without an explicit health or Garmin context.

Medium
Category
Trigger Abuse
Confidence
95% confidence
Finding

L287 instructs the agent to fetch summary data when asked “how am I doing?” or similar, a phrase that also commonly concerns work, learning, or other unrelated topics. Although the guide is Garmin-specific, this activation instruction does not require an established Garmin context or an explicit request to analyze health data.

Content

No source excerpt is available for this finding.

The profile command exposes account identity information beyond the stated health-analysis purpose.

Low
Category
Excessive Agency
Confidence
85% confidence
Finding

The manifest describes retrieving and analyzing health metrics, workouts, and routes. This function instead retrieves the user's full name and includes account display-name and email fields, when present, in the output; exporting an account email is not needed for the described analyses.

Content

No source excerpt is available for this finding.