Back to skill

Security audit

weixin-quote

Security checks across malware telemetry and agentic risk

Overview

This skill is purpose-aligned, but it asks an agent to replace the official WeChat plugin with an unpinned third-party fork that will handle message context.

Install only if you are comfortable replacing the official WeChat plugin with this publisher's fork. Prefer a pinned and reviewed package version, confirm the publisher and source, understand that quoted message text will be processed for context injection, and keep a rollback path to the official plugin.

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

Error
Location
SKILL.md:14
Finding
Official WeChat Plugin Replaced with an Unpinned Third-Party Fork## Vulnerability Details **File Location**: `SKILL.md`, lines 14-17 **Vulnerability Type**: Unpinned third-party dependency and unsafe supply-chain replacement **Risk Level**: High ### 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`):先 `openclaw plugins uninstall openclaw-weixin`,再继续。 - 若未安装:`openclaw plugins install clawhub:@yechang1450/openclaw-weixin-clawbot`(内含引用 id 匹配与全文注入的微信渠道插件;请先征得用户同意)。 ``` The same replacement behavior is also documented in `README.md`, lines 9-12, and `README-zh.md`, lines 9-11. ### Technical Analysis The skill directs the agent to remove the official WeChat channel plugin and install the publisher-scoped third-party package `clawhub:@yechang1450/openclaw-weixin-clawbot`. The dependency is not pinned to an immutable version or integrity digest. Its source code is not included in the audited project, which contains only documentation and instructions. Consequently, the effective implementation loaded by the gateway may change after this skill has been reviewed. The fork operates in a sensitive message-processing position: according to the project documentation, it records server message identifiers, resolves quoted messages, and injects complete quoted message text into model context. A malicious or compromised dependency release could therefore process private chat content or execute other actions available to the OpenClaw plugin runtime. User consent before package installation is a useful safeguard against accidental replacement, but it does not verify package integrity, provenance, runtime permissions, data retention, or network behavior. ### Attack Path 1. An attacker compromises the third-party publisher account, package registry entry, release pipeline, or distribution source for `@yechang1450/openclaw-weixi ...[truncated 1783 chars]
Remediation
## Remediation Suggestions 1. Pin the dependency to a specific, reviewed version rather than installing a mutable package identifier. 2. Verify a cryptographic digest or signed provenance record before installation. 3. Publish or vendor the exact plugin source corresponding to the pinned release so its message handling, network behavior, storage, and command execution can be audited. 4. Verify the package publisher, registry namespace, build pipeline, and release signatures through an independent trust channel. 5. Request explicit informed consent immediately before both uninstalling the official plugin and installing the fork. Explain that the replacement processes message content. 6. Avoid removing the official plugin until the replacement has passed provenance and integrity checks. Provide a documented rollback procedure. 7. Run the plugin with least privilege, restricting filesystem access, environment variables, credentials, subprocess execution, and outbound network access to what is strictly required. 8. Document how message identifiers and quoted text are stored, retained, logged, and deleted. Avoid logging full message content. 9. Add automated checks that reject unexpected package versions, sources, signatures, or integrity hashes before restarting the gateway. 10. Prefer contributing the required quote-handling functionality to the official plugin or using an officially reviewed extension mechanism where possible.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (2)

Vague Triggers

Low
Confidence
84% confidence
Finding
The README lists activation phrases such as “启用引用”, “引用没生效”, and especially “quote” without clarifying the exact context in which the skill should or should not activate. Because the trigger section does not provide exclusion conditions or narrower invocation boundaries, these phrases could overlap with ordinary support or conversational requests and cause unintended invocation.

Scope Creep

Low
Category
Excessive Agency
Content
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
Confidence
70% confidence
Finding
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.