Back to skill

Security audit

Hourly Knowledge

Security checks for vulnerabilities and agentic risk

Overview

The skill is not malicious, but it needs review because it sends recurring automated messages and its multi-user isolation claim is not supported by the included history storage script.

Install only if you want automatic hourly knowledge messages and can manage consent/opt-out outside the skill. For multi-user use, require per-user or per-channel history storage before deployment so one user's history cannot influence another user's content selection.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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
Findings (7)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The documented behavior claims hourly push, search, generation, and multi-user isolation, but the described implementation relies on a single global history file and omits concrete controls for per-user separation. This mismatch is dangerous because operators may deploy the skill under false assumptions, leading to cross-user data mixing, unintended behavior, or exposing internal state through undocumented interfaces.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
83% confidence
Finding

The skill instructs the agent to read and update a history file but does not declare any tool scope or permissions boundaries. In practice, this can cause the runtime to grant broader-than-necessary file access or make the skill's write behavior invisible to reviewers, increasing the chance of unauthorized file modification if the skill is changed or abused.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill is designed to automatically push content to users on a schedule, but the description lacks a clear user-facing warning or consent model for recurring outbound messages. This can result in unsolicited messaging, surprise notification behavior, and misuse in environments where automated pushes require explicit disclosure or opt-in.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The script uses a single global history file under a shared workspace path for all invocations, while the skill claims to support multiple users. In a multi-user deployment, one user's pushed topics can influence or suppress another user's content selection, causing cross-user data leakage and integrity issues in personalization/deduplication behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The natural-language instructions and example output are entirely in Chinese, and the skill does not state that the locale is optional, user-selected, or region-specific. This can violate language/locale policy expectations when a skill implicitly forces one language without user opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The file’s natural-language documentation is written only in Chinese and indicates the skill’s behavior in that locale without any visible option for alternate language use. Under the stated policy, forcing a specific language without user opt-in can be a locale-policy issue unless the regional constraint is explicitly documented and justified.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
98% confidence
Finding

The function named and documented as returning blocked topic keywords actually returns the full topic history unchanged. This can cause policy logic to operate on the wrong data set, leading to overblocking, underblocking, or accidental reuse of user-derived history as a moderation/control list; in this skill context it mainly degrades correctness and may expose prior topic data through interfaces expecting only keywords.

Content

No source excerpt is available for this finding.