ClawHub
Back to skill

Security audit

CareMax Members

Security checks for vulnerabilities and agentic risk

Overview

The skill matches its family health purpose, but it relies on unpinned external auth scripts and can start a background login flow for sensitive health data without clear user confirmation.

Install only if you trust the CareMax skill source and have reviewed or pinned the caremax-auth dependency. Before use, confirm which CareMax account, member profile, and server URL will be accessed, and avoid background authentication or custom URLs unless you understand where credentials and health data will go.

A.I.G

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:10
Finding

Unpinned Third-Party Installation and Execution of Unaudited Dependency Scripts

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 10–20
Vulnerability Type: T08: Insecure Dependencies
Risk Level: Medium

Vulnerable Code

markdown
> **Requires `caremax-auth` as a sibling directory** (`../caremax-auth/`). If missing, tell the user to install caremax-auth first (e.g. `npx skills add KittenYang/caremax-skills`).

List and work with family member profiles. CareMax supports tracking health data for multiple family members.

## Prerequisites — Auto-Auth (MANDATORY)

```bash
APICALL="bash ../caremax-auth/scripts/api-call.sh"

If api-call.sh returns {"error":"no_credentials",...} → immediately run bash ../caremax-auth/scripts/auth-flow.sh [base_url] in background (from this skill’s root). If the user specified a custom URL, pass it as the argument.

text

### Technical Analysis

The Skill recommends installing `KittenYang/caremax-skills` through `npx` without specifying an immutable package version, commit hash, checksum, or other integrity control. It subsequently mandates execution of `api-call.sh` and `auth-flow.sh` from the separately installed `../caremax-auth/` directory.

Because the referenced scripts are not included in the audited project, their implementation, credential handling, command construction, network destinations, and update behavior cannot be verified. The effective behavior of this Skill therefore depends on mutable external supply-chain content. A compromised package source, repository, release, or dependency could cause attacker-controlled code to be installed and executed with the permissions of the Agent process.

The authentication flow also accepts a user-specified `base_url`. Validation of that value cannot be confirmed because the receiving script is absent. This increases the importance of auditing the dependency and enforcing an allowlist of trusted authentication endpoints.

### Attack Path

1. A user invokes the family-member Skill while 
...[truncated 1618 chars]
Remediation
View remediation

Remediation Suggestions

  1. Pin the installer and CareMax dependency to reviewed, immutable versions or repository commit hashes rather than relying on a mutable source.
  2. Verify downloaded content using cryptographic checksums or signed release metadata before executing it.
  3. Include reviewed authentication scripts in the audited package where practical, or audit and version them together with this Skill.
  4. Require explicit user confirmation before installing external dependencies or starting an authentication process.
  5. Restrict base_url to an allowlist of trusted HTTPS origins and reject malformed URLs, untrusted hosts, redirects to unapproved origins, and plaintext HTTP.
  6. Run authentication and API scripts with least privilege in a sandbox that limits filesystem, environment-variable, credential, and network access.
  7. Document the exact dependency version, expected script hashes, permitted network endpoints, and credential-storage behavior.
  8. Re-audit api-call.sh, auth-flow.sh, and their transitive dependencies before treating the Skill as trusted.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill is explicitly designed to access spouse, child, parent, and other family members’ health data, but it provides no requirement to verify authorization, consent, or user intent before retrieving sensitive records. In a health-data context, this is especially dangerous because it can enable unauthorized disclosure of highly sensitive personal and medical information across linked profiles.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
89% confidence
Finding

The skill instructs users to install or invoke tooling via npx skills without pinning a specific version or immutable source. That creates a supply-chain risk: a malicious or compromised package version could be fetched and executed at runtime, and the skill’s healthcare context increases sensitivity because any compromise could expose protected health data or credentials.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill instructs the agent to automatically launch a background authentication flow immediately upon missing credentials, without user disclosure or consent. Silent auth initiation is risky because it can trigger unexpected credential handling, browser/device prompts, or token issuance flows, and in this healthcare setting it may lead to unauthorized session creation or access to protected medical data.

Content

No source excerpt is available for this finding.