ClawHub

Clawcrm

Security checks across malware telemetry and agentic risk

Overview

ClawCRM is a coherent CRM integration, but it gives an agent broad authority to enrich contacts, change CRM data, and send delayed sales emails without clear approval, consent, or cancellation safeguards.

Install only if you are comfortable giving an agent CRM administrator-like authority. Use a dedicated scoped key if available, test in a sandbox, require explicit human approval before lead enrichment or any outbound email sequence, and confirm the service’s consent, opt-out, retention, deletion, rate-limit, audit-log, and cancellation controls before using it with real contacts.

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 (3)

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill explicitly promotes autonomous lead enrichment, proposal generation, engagement tracking, and email sequencing but does not warn about consent, lawful basis, recipient expectations, or handling of personal/business data. In an agent context, this increases the chance that an AI system will process third-party personal data and initiate outreach without human review, creating privacy, compliance, and reputational risk.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The documentation returns and then reuses a highly privileged admin token in headers without any warning to treat it as a secret, avoid logging it, or scope/rotate it. Because the skill is designed for agent automation, such examples can lead to tokens being embedded in prompts, scripts, logs, or shared traces, enabling unauthorized CRM access and bulk actions if exposed.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The email-sequence feature documents autonomous outbound messaging with templated personalization and delayed follow-ups but does not include safeguards such as recipient verification, consent checks, rate limits, human approval, or anti-spam warnings. In an AI-agent workflow, this can facilitate unintended or large-scale outreach to real contacts, causing spam, policy violations, and brand damage.

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal