Back to skill

Security audit

postey-teams

Security checks for vulnerabilities and agentic risk

Overview

This skill supports reviewing social posts and explicitly requires your instruction before publishing or scheduling.

Use a trusted Postey integration and remember that anyone holding a share link can read its draft. Reviewer approval does not authorize publication. This review covers the supplied skill files, not the separate Postey integration.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 65)May include surrounding context.

md
| What is true | What it means for you | The failure it prevents |
|---|---|---|
| A draft carries no approval flag | Approval lives in an internal comment, a message, or a conversation. Read it; never infer it. | Treating a status as consent |
| Silence is not approval | If asked whether something is cleared, say who said what and when — never "no objections were raised" | Reporting absence of dissent as sign-off |
| Nothing stops a `DRAFT` publishing | The status does not protect it. Only the rule that publishing needs explicit instruction does, and that rule lives in the agent, not the server. | Assuming the server will catch you |