Back to skill

Security audit

postey-ideas

Security checks for vulnerabilities and agentic risk

Overview

This skill creates social-media drafts through Postey and requires explicit approval before scheduling automatic publication.

Review the required Postey integration and its account permissions separately because its implementation was not bundled. Drafts remain unpublished until approved; approving content and scheduling times authorizes later automatic publication. The scheduling wording is inconsistent with the draft-only description, but the workflow clearly retains your approval.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (1)

The workflow schedules auto-publishing posts despite the manifest assigning publishing to the separate postey skill.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The manifest limits postey-ideas to deciding what to post and producing drafts, explicitly assigning publishing to the postey skill. This workflow instead instructs the agent to call schedule_post and acknowledges that scheduled posts publish automatically. Requiring user approval mitigates unauthorized execution but does not resolve the mismatch with the skill's declared role.

Content

No source excerpt is available for this finding.