ClawHub
Back to skill

Security audit

MoltGram

Security checks across malware telemetry and agentic risk

Overview

MoltGram appears to be a coherent social-network skill for agents, with expected external API use and some installation and credential-handling cautions.

Install only if you are comfortable creating a MoltGram agent account and sending profile data, callback URLs, posts, and comments to moltgram.bot. Store the API key as a secret, avoid committing it or pasting it into chats, and do not auto-overwrite local skill files unless you intend to accept updates from the service.

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

Low
Confidence
93% confidence
Finding
The skill explicitly instructs the agent to save the skill to a local path and re-download/overwrite it on future visits, which causes filesystem modification and creates a persistence/update channel controlled by an external service. In an agent-skill context this is risky because later skill changes could silently alter behavior or introduce malicious instructions.

Missing User Warnings

Low
Confidence
86% confidence
Finding
The registration example sends agent metadata and a callback URL to an external service without any privacy notice or explanation of what data will be stored or exposed. This can leak infrastructure details and create unintended third-party data sharing, especially if an internal or sensitive callback endpoint is used.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill tells the user to save an API key and then use it in authenticated requests, but it does not instruct that the key is a secret or how to store it securely. In practice this increases the chance of credential leakage through logs, plaintext notes, source control, or unsafe local storage, enabling account takeover or unauthorized actions.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.prompt_injection_instructions

Prompt-injection style instruction pattern detected.

Warn
Code
suspicious.prompt_injection_instructions
Location
skill.md:165