ClawHub
Back to skill

Security audit

Skillboss

Security checks across malware telemetry and agentic risk

Overview

SkillBoss is a coherent remote API gateway, but it gives agents broad live email, SMS, OTP, document, and audio-processing powers without enough consent or privacy guardrails.

Install only if you trust SkillBoss as a broad remote gateway and can control use of the API key. Use restricted or budget-limited credentials where possible, avoid confidential documents or recordings unless authorized, and require explicit human review before any email, SMS, OTP, or batch communication is sent.

SkillSpector

By NVIDIA
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 (11)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill documents email and SMS/OTP actions that send recipient addresses, phone numbers, message contents, and verification data to an external API, but it does not warn the user that this causes third-party transmission of potentially sensitive personal data. In an agent setting, this can lead to privacy violations or unintended outbound communications because the skill normalizes high-impact actions without disclosure or consent guidance.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The document processing section instructs users to provide a document URL to an external parsing/extraction service without warning that the document contents may be fetched, uploaded, and processed by third parties. If agents use this on confidential files, sensitive business or personal data could be exposed outside the local trust boundary.

Missing User Warnings

Low
Confidence
88% confidence
Finding
The image generation example downloads a remote URL returned by the service and writes it to a local file, but it does not warn that this modifies the filesystem and trusts remote content. In agent workflows, silent file writes can overwrite data, store untrusted content locally, or create artifacts without user awareness.

External Transmission

Medium
Category
Data Exfiltration
Content
## Speech-to-Text

```bash
curl -s -X POST https://api.heybossai.com/v1/run \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
79% confidence
Finding
This duplicate hit corresponds to the STT example that uploads raw audio externally. Because voice data is often sensitive and the documentation lacks a privacy warning, the context makes this more dangerous than ordinary API traffic.

External Transmission

Medium
Category
Data Exfiltration
Content
## Email

```bash
curl -s -X POST https://api.heybossai.com/v1/run \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
94% confidence
Finding
This duplicate email endpoint hit corresponds to a real outbound communication capability with clear external side effects. In a skill for autonomous agents, lack of consent/authorization guidance makes this materially risky even though the author's likely intent is benign.

External Transmission

Medium
Category
Data Exfiltration
Content
Save to file:

```bash
URL=$(curl -s -X POST https://api.heybossai.com/v1/run \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "mm/img", "inputs": {"prompt": "A sunset over mountains"}}' \
Confidence
76% confidence
Finding
This example not only calls the external API but then downloads a returned URL and saves it locally. While not malicious, in context it combines external transmission with local side effects and trust of remote content, which warrants a warning in an agent-executed skill.

External Transmission

Medium
Category
Data Exfiltration
Content
## Speech-to-Text

```bash
curl -s -X POST https://api.heybossai.com/v1/run \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
79% confidence
Finding
This duplicate hit corresponds to the STT example that uploads raw audio externally. Because voice data is often sensitive and the documentation lacks a privacy warning, the context makes this more dangerous than ordinary API traffic.

External Transmission

Medium
Category
Data Exfiltration
Content
## Document Processing

```bash
curl -s -X POST https://api.heybossai.com/v1/run \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
91% confidence
Finding
The document-processing request sends a document URL to an external service for parsing/extraction, which may expose sensitive file contents to third parties without any accompanying warning. In an agent context, this can easily lead to unintended disclosure of internal PDFs or DOCX files.

External Transmission

Medium
Category
Data Exfiltration
Content
## Email

```bash
curl -s -X POST https://api.heybossai.com/v1/run \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
94% confidence
Finding
This duplicate email endpoint hit corresponds to a real outbound communication capability with clear external side effects. In a skill for autonomous agents, lack of consent/authorization guidance makes this materially risky even though the author's likely intent is benign.

External Transmission

Medium
Category
Data Exfiltration
Content
Send OTP:

```bash
curl -s -X POST https://api.heybossai.com/v1/run \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
93% confidence
Finding
Sending OTPs transmits phone numbers to an external verification provider and causes real-world side effects, yet the skill does not warn about privacy implications or the need for authorization. This can enable unsolicited messages, disclosure of personal data, or abuse of verification flows in agent-driven contexts.

External Transmission

Medium
Category
Data Exfiltration
Content
Verify OTP:

```bash
curl -s -X POST https://api.heybossai.com/v1/run \
  -H "Authorization: Bearer $SKILLBOSS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
Confidence
90% confidence
Finding
OTP verification sends phone numbers and verification codes to an external service, involving sensitive authentication-related data. Without warnings or guidance, an agent may mishandle these flows or expose authentication metadata to third parties unexpectedly.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.