T09 · Insecure Skill Coding Practices
- Location
SKILL.md:43- Finding
Plaintext Storage of WeCom Application Secrets and Webhook Credentials
- Content
View full analysis
- Remediation
View remediation
Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent WeCom integration guide, but it asks users to configure powerful messaging credentials and customer-message automation without enough security or privacy safeguards.
Install only if you are authorized to connect OpenClaw to your organization's WeCom tenant. Treat app secrets, access tokens, and webhook URLs as credentials: restrict config file permissions, do not commit or share them, redact them from logs and screenshots, and rotate them if exposed. Confirm your organization permits automated monitoring and replies for customer messages before enabling that workflow.
SKILL.md:43Plaintext Storage of WeCom Application Secrets and Webhook Credentials
The skill instructs users to place sensitive WeCom credentials such as CorpId, AgentId, and especially the application Secret directly into a local plaintext config file, but provides no warning about secure storage, file permissions, redaction, or accidental disclosure through backups and logs. In an integration skill, this context makes the issue more dangerous because these credentials enable message sending and API access if the host or config is exposed.
The skill promotes automatic monitoring and replying to customer WeCom messages without any discussion of privacy, consent, data retention, or organizational policy controls. Because this involves customer communications, the omission can lead to unauthorized processing of personal or sensitive business data and noncompliant automation.
The configuration examples include group bot webhook URLs without warning that the webhook key embedded in the URL is effectively a bearer secret. If copied into tickets, repos, chat, or screenshots, an attacker could abuse the webhook to send spoofed messages into enterprise group chats.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
curl -X POST "YOUR_WEBHOOK_URL" \
-H "Content-Type: application/json" \
-d '{"msgtype":"text","text":{"content":"测试"}}'
The description and the rest of the skill content are presented only in Chinese, and the author metadata frames it as a Chinese ecosystem skill. Under the policy rule, forcing a specific language without user choice can be a natural-language policy violation unless the locale restriction is explicitly justified.
The description and keywords indicate the skill is targeted specifically at Chinese/WeCom usage (e.g. Chinese description text and the keyword "chinese"), but the file does not state that this locale restriction is optional or justified as a region-specific tool. Under the policy rule, forcing or implying a specific language/locale without opt-in can be a natural-language policy concern.