Back to skill

Security audit

guild-knowledge

Security checks for vulnerabilities and agentic risk

Overview

The skill is not malicious, but it can activate on ordinary requests and automatically read local Guild documents and search the web, so it needs review before installation.

Install only if you are comfortable with the skill activating automatically on broad phrases and potentially sending task-related search queries externally. Prefer narrowing or disabling auto-triggers, using explicit /guild commands, and requiring confirmation before web searches or any Guild document changes.

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
  • 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 (7)

Vague Triggers

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The auto-trigger list includes broad, common requests such as 'build website', 'cleanup workspace', and 'delete files', which can cause the skill to activate during routine conversations without clear user intent. Because the skill then mandates document reads and web searches, this creates a real risk of unintended external actions, privacy leakage, and unsafe influence over file-management tasks.

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill is configured with auto_trigger=true and broad keywords such as 'build website', 'delete files', 'organize files', and 'new technology' that commonly appear in normal conversations. This can cause the skill to activate without clear user intent, leading to unsolicited document reads, searches, and workflow injection into unrelated tasks.

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

The activation condition 'user mentions a related task' is too subjective and does not provide a reliable boundary for invocation. In practice, this can make the skill engage on many ordinary requests, causing unintended retrieval, search, and recommendation behavior that may expose context or alter agent behavior without explicit consent.

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill is configured to auto-trigger on broad, everyday phrases such as 'build website', 'delete files', 'organize files', and 'new technology'. Because the skill also advertises actions like auto-reading documents and searching/comparing information, these generic triggers can cause unintended invocation in unrelated conversations, increasing the chance of unexpected data access, workflow interference, or context-sensitive actions being initiated without clear user intent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The overview and feature description state that the skill automatically reads documents and performs mandatory latest-information searches, but they do not prominently warn users that external web searches may occur automatically before tasks. This can surprise users, expose sensitive task context to search providers, and undermine informed consent, especially when the trigger mechanism is already broad.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

Allowing users to customize trigger keywords without any scope limits, safety rules, or examples of prohibited broad terms makes activation behavior unpredictable. This increases the chance that the skill will be bound to generic phrases and fire in contexts where the user did not intend to invoke Guild workflows.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
63% confidence
Finding

The manifest declares supported languages as only zh-CN and en, but does not state that users can choose among them or explain any locale limitation. For a natural-language skill, hard-coded language constraints without opt-in or justification can violate language/locale policy expectations.

Content

No source excerpt is available for this finding.