Security audit

Alpha Feed Creator

Security checks for vulnerabilities and agentic risk

Overview

The skill is not clearly malicious, but it asks for recurring automated collection, shared file writes, and group-message pushes without enough scoping or lifecycle controls.

Review before installing. Use a dedicated low-privilege account, keep API keys and push tokens in a secret manager, run dry-run tests first, verify all push targets and shared output paths, and avoid enabling cron or administrator execution unless you explicitly need recurring unattended delivery.

A.I.G

Vulnerability Patterns

System PersistenceInstalls backdoors, hooks, services, or scheduled tasks that survive the run
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

Findings (1)

T06 · System Persistence

Error

Location: SKILL.md:174
Finding: Persistent Unattended Execution Through Scheduled Task Registration
Content: View full analysis
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (3)

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The activation language is overly broad and encourages use for generic efficiency, automation, and workflow optimization tasks rather than narrowly scoped content collection. In agent environments, this increases the chance the skill is auto-invoked for unrelated requests, potentially granting it access to exec, read, glob, and grep capabilities in contexts where the user did not intend data collection, file access, or outbound pushing.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The skill advertises automatic group pushing and writing reports to shared storage, but does not present a prominent warning, consent gate, or data-classification guidance before disclosure occurs. Because the skill aggregates content across sources and supports multi-tenant/team contexts, accidental inclusion of sensitive data, internal research, API-derived content, or customer information could be persisted or broadcast to unintended recipients.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The documented return schema is inconsistent with the skill’s stated purpose: it emits a generic evaluation/scoring report rather than structured content-collection results. This mismatch can cause downstream agents or users to mis-handle outputs, apply the skill in the wrong workflow, or trust fabricated success signals instead of verifying what data was collected and where it was sent or stored.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.