ClawHub

Tweet Composer

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only tweet-writing advisor that may encourage reach optimization but does not access accounts, post content, or install executable code.

Install this only if you want an advisory assistant for drafting and scoring tweets. Treat claims about X's exact ranking behavior as interpretive rather than authoritative, and avoid using the reach-optimization advice for spam, engagement bait, or platform-policy violations.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
96% confidence
Finding
The README includes very broad invocation examples such as 'Optimize this thread for maximum reach' and 'Should I post this now or wait for peak hours?' that resemble normal user conversation rather than a clearly namespaced command. In agent ecosystems where skills are auto-selected from conversational text, this can cause unintended activation, leading the agent to route unrelated user input into this skill and potentially expose drafts, strategy discussions, or other context to the skill unexpectedly.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The skill advertises a broad set of activation scenarios such as general posting strategy advice, which can cause the agent to invoke this skill for loosely related social-media requests. That increases the chance of over-triggering specialized guidance when the user did not explicitly ask for tweet optimization, leading to irrelevant behavior, unintended file reads, or policy-bypassing context injection from the skill.

Missing User Warnings

Low
Confidence
88% confidence
Finding
The flagged section gives concrete tactics for maximizing engagement based on a platform ranking system, including behavioral nudges intended to increase replies, shares, dwell time, and profile clicks. In this skill’s context, that moves beyond neutral education into operational guidance for gaming feed distribution, which can facilitate spammy or manipulative posting behavior at scale even if it is not overtly malicious.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal