ClawHub
Back to skill

Security audit

Social Posting

Security checks for vulnerabilities and agentic risk

Overview

This is a documented social-media posting skill whose sensitive actions are expected for its purpose, but users should configure credentials and posting confirmations carefully.

Install only if you intend to let an agent help connect social accounts and publish or schedule posts through PostForMe or LATE. Confirm the exact account, platforms, content, media, and schedule before each posting action, and avoid global fallback credentials unless an administrator has intentionally scoped and audited them. Verify that encrypted credentials and post history can be reviewed, revoked, and deleted.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger description is broad enough to activate on common phrases like 'social post' or 'publish to platforms,' which increases the chance the skill runs in contexts where the user did not explicitly intend live posting. Because this skill can publish to real social accounts across multiple platforms, accidental invocation can cause unintended external actions and reputational harm.

Missing User Warnings

Medium
Confidence
98% confidence
Finding
The description omits a clear warning that the skill can publish content to live accounts and may fall back to global environment-variable credentials if user credentials are unavailable. That combination creates a significant risk of users not understanding that real external posting will occur, potentially under non-user-scoped credentials, leading to unauthorized or surprising publication.

Static analysis

No suspicious patterns detected.