ClawHub

Explorium AgentSource

Security checks across malware telemetry and agentic risk

Overview

This is a coherent Explorium prospecting skill, but it should be reviewed because it can expose an API key in terminal output and leaves sensitive prospecting results in predictable shared temp files.

Install only if you intend to use Explorium AgentSource and are comfortable with prospect/contact enrichment. Do not enter the API key through setup on shared, logged, or screen-shared terminals unless the script is fixed not to echo or reprint it. Treat /tmp/agentsource_*.json as sensitive, delete those files after use, and confirm that any personal email or phone enrichment complies with your legal, privacy, and outreach obligations.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
98% confidence
Finding
The script echoes a shell export command containing the user-supplied API key, which unnecessarily displays the secret on screen and may expose it via terminal scrollback, screen sharing, logging, or shell session capture. This is a real secret-handling issue because the key is revealed after entry even though setup does not require re-printing it.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The CLI explicitly writes all API results to /tmp using predictable filenames based only on timestamp and command. In this skill's context, those results can contain sensitive business intelligence, prospect/contact data, and enrichment outputs, so storing them in a shared world-accessible temp location materially increases the risk of local disclosure or tampering by other users/processes on the same host.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The documentation explicitly promotes enrichment of prospects with personal email addresses, direct phone numbers, and mobile numbers, but provides no privacy, consent, lawful-basis, or data-handling guidance. In a B2B prospecting skill, this omission is materially risky because it normalizes collection and export of sensitive contact data for outreach, increasing the chance of non-compliant use under privacy, marketing, and data-protection regimes.

Session Persistence

Medium
Category
Rogue Agent
Content
## CLI Reference

All commands write results to `/tmp/agentsource_<timestamp>_<command>.json` and print only the file path to stdout.

### API Key Config
```bash
Confidence
83% confidence
Finding
write results to `/tmp/agentsource_<timestamp>_<command>.json` and print only the file path to stdout. ### API Key Config ```bash python3 ~/.agentsource/bin/agentsource.py config --api-key <key> ```

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal