Back to skill

Security audit

weixin-debounce20s

Security checks across malware telemetry and agentic risk

Overview

The skill is openly designed to install a WeChat debounce feature, but it does so by replacing the official WeChat plugin with an unpinned third-party fork and restarting the gateway.

Install only if you intentionally want to replace the official WeChat channel with this fork. Before running it, confirm the exact plugin package/version, understand that the gateway will restart, and keep a rollback plan for reinstalling the official plugin and restoring prior OpenClaw configuration.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:17
Finding
Unpinned Third-Party Plugin Replaces the Official WeChat Plugin## Vulnerability Details **File Location**: `SKILL.md`, lines 17-20 **Vulnerability Type**: `T08: Insecure Dependencies` **Risk Level**: Medium ### Vulnerable Code Snippet ```markdown 1. **判断微信渠道插件状态** - 运行 `openclaw plugins inspect openclaw-weixin --json`。 - 若已存在且 source 含 `clawbot-plugins` 或显示为 ClawBot fork 包:跳第 2 步。 - 若已装官方版(`@tencent-weixin/openclaw-weixin`,source 非 fork):先 `openclaw plugins uninstall openclaw-weixin`(配置与登录态保留在 `.openclaw`,不受影响),再继续。 - 若未安装:`openclaw plugins install clawhub:@yechang1450/openclaw-weixin-clawbot`(内含 20s 防抖 + 引用注入的微信渠道插件;请先征得用户同意再执行安装)。 ``` ### Technical Analysis The Skill directs the agent to uninstall the official WeChat plugin and install `clawhub:@yechang1450/openclaw-weixin-clawbot`, a third-party fork, without specifying an immutable version or integrity digest. Package consent informs the user of the installation but does not verify the package's provenance or contents. The third-party plugin's source code is not included in the audited project. Consequently, its claimed debounce and quote-injection behavior, its requested permissions, and its handling of credentials and message data cannot be verified from this repository. Because the dependency is unpinned, the code resolved during a future installation may differ from the code reviewed or expected by the Skill author. ### Attack Path 1. A user invokes the Skill to enable WeChat message debouncing. 2. The agent inspects the installed plugin and detects the official package or no existing package. 3. If present, the agent removes the official WeChat plugin. 4. After obtaining installation consent, the agent resolves and installs the unpinned third-party ClawHub package. 5. An attacker who has compromised the publisher account, package registry, package distribution process, or a later package release supplies modified plugin code. 6. The Skill restarts the OpenClaw gateway, causing the downloaded plugin code t ...[truncated 920 chars]
Remediation
## Remediation Suggestions 1. Pin the third-party plugin to a specific, reviewed version rather than resolving the latest mutable release. 2. Verify the package with a cryptographic integrity digest or signed provenance metadata before installation. 3. Publish or reference the exact auditable source commit corresponding to the pinned package artifact. 4. Review the plugin's installation hooks, runtime behavior, requested permissions, credential access, message handling, and outbound network activity. 5. Prefer the official plugin when the required behavior can be implemented through supported configuration or an independently reviewable minimal extension. 6. Clearly disclose that the official plugin will be replaced, identify the third-party publisher, and explain the plugin's effective privileges before requesting consent. 7. Run the plugin under least privilege, restricting filesystem, credential, process, and network access to only what the debounce feature requires. 8. Preserve a verified rollback path to reinstall the official plugin if package verification, installation, or runtime validation fails.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The trigger phrase “连发合并” is broad and plausibly appears in normal conversation about WeChat messaging behavior, troubleshooting, or feature discussion. Because this is a setup skill that can install/replace plugins and restart the gateway, accidental invocation could cause unintended configuration changes on the host, making the overlap materially risky in context.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The README explicitly states that running the skill can replace the installed WeChat plugin, modify configuration, restart the gateway, and perform verification steps, but it does not clearly foreground the operational impact, rollback considerations, or trust implications of switching to a forked plugin. In a setup/enabler skill, this omission is security-relevant because users may trigger the skill expecting a harmless fix while it performs privileged system changes and installs third-party code.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.