Security audit

Discuz Admin Assistant

Security checks for vulnerabilities and agentic risk

Overview

This is a Chinese Discuz! admin reference skill with no executable installer, but it covers powerful site-admin actions that require care.

Install this only if you want a Chinese-language Discuz! administrator reference. Treat any advice involving deletion, restores, SQL, templates, third-party scripts, APIs, payment settings, or credentials as admin-only work: verify backups, restrict access, use trusted sources, and test changes before applying them to a live site.

A.I.G

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

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (17)

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 94% confidence
Finding: These instructions describe a full data restore that overwrites existing site data and directs the operator to place a restore.php script into a web-accessible data directory, then invoke it via browser. In an admin-assistant skill, surfacing these steps without strong safety gating increases the risk of accidental destructive recovery actions and temporary exposure of a sensitive restore endpoint if the script is not removed promptly.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 88% confidence
Finding: This markdown file includes instructions for deleting filtered posts, topics, and users in bulk, including deleting associated UCenter profile data. While some backup advice appears elsewhere, these deletion procedures themselves do not clearly warn the reader about irreversible data loss at the point of action.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The RESTful API section instructs administrators to enable external API access and import official or third-party XML interface definitions without any trust-validation or exposure warning. In an admin-assistant skill, this can normalize importing untrusted extensions into a privileged environment, potentially expanding attack surface, exposing sensitive forum data, or enabling insecure endpoints and excessive privileges.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: This markdown file is entirely written in Chinese and provides no indication that the user can choose another language or locale. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale constraint is explicitly documented and justified.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 83% confidence
Finding: This markdown file documents a merge action where the source forum's posts are moved and the source forum is deleted. Although the consequence is mentioned, the skill text does not present it as a prominent caution or warning to the user before the action, which is important for an irreversible administrative change.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 88% confidence
Finding: This markdown file documents a destructive action: deleting a region entry directly from settings. The description explains how to delete it but provides no warning about irreversibility, downstream data impact, or need to confirm before removal.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: This markdown file includes instructions to delete a board or partition by directly clicking a delete link, but it does not explicitly warn that the action is destructive and may permanently remove associated structure or content. For markdown files, safety-impacting behaviors should include clear warnings about effects on user data or system integrity.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The documentation explicitly tells administrators to paste third-party statistics code into the site, which commonly means arbitrary JavaScript executes in every visitor's browser. Even though this is a normal admin feature, omitting any warning about script trust, privacy disclosure, and supply-chain risk can lead to XSS-like behavior, user tracking, and compromise if the injected code is malicious or the provider is later breached.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The guide states that arbitrary HTML can be inserted into the page head, which can enable script injection, remote resource loading, malicious meta refresh, and broad client-side compromise across the site. In an admin-assistant context this is more dangerous because it normalizes unrestricted code insertion into a globally rendered template area without any security caveat.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The documentation describes connecting to third-party module URLs, exchanging client IDs and communication keys, and consuming external XML data without warning about trust boundaries or integrity validation. This creates supply-chain and SSRF-like risk, can expose secrets, and may import malicious content or unsafe remote dependencies into the portal rendering path.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: This markdown file presents all user-facing instructions and descriptions exclusively in Chinese, and there is no indication that the skill is region-specific or that users can opt into another language. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 78% confidence
Finding: This markdown file discusses configuring a UCenter communication key and database username/password, but it does not explicitly warn users to protect these secrets from disclosure or avoid sharing them insecurely. Because the file handles credentials in a user-facing setup guide, a clear warning about secrecy and secure storage would improve safety.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 78% confidence
Finding: This markdown file instructs users that they can delete circles and merge categories, and notes that the original category is deleted after merging. Under the markdown-specific warning criterion, destructive actions affecting user data/system content should be accompanied by an explicit warning about impact or irreversibility, which is not clearly present here.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 90% confidence
Finding: A language or locale policy violation applies to natural-language content across all file types when a specific language is forced without user opt-in. This file presents all instructions in Chinese and does not indicate that the skill is region-specific or provide an alternative language option.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 81% confidence
Finding: This markdown file describes modifying navigation and theme list templates, which can affect site behavior and integrity. Although it says users should have some code knowledge and avoid certain code types, it does not explicitly disclose the risk of breaking pages or causing site instability when making template changes.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 71% confidence
Finding: L445 states that topic cover images may be provided via local upload, which is a data-affecting operation, but the description gives no caution about validating file sources or handling uploaded content safely. While common, the file operation is presented without any user-facing warning.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The entire markdown file is written in Chinese and presents user-facing instructional content with no indication that other languages or locale options are available. Under the policy rule for language or locale constraints, this is a natural-language policy concern because the file imposes a single language without documented opt-in or justification.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.