Security audit
Security checks for vulnerabilities and agentic risk
This skill is clearly for creating WeChat Official Account drafts, but it can modify a live account draft box and relies on missing/uninspected helper scripts without a clear confirmation gate.
Install only if you intend agents to create or change drafts and upload images in a real WeChat Official Account. Require the agent to summarize the target account, title, content, and uploads and get your explicit approval before any API call; also verify the missing helper scripts from a trusted source before using credentials.
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
./scripts/get_access_token.sh
The trigger phrase list includes broad terms such as “微信文章” and “公众号发文”, which can match ordinary writing requests and cause the skill to activate in situations where the user did not clearly intend draft-box publication. In this skill, overbroad activation is more dangerous because the workflow culminates in creating or modifying remote WeChat draft content, not just local text generation.
The skill instructs direct saving of generated content to a live WeChat public-account draft box without an explicit warning that it will modify remote account data. This creates a risk of unintended external side effects, including publishing workflow disruption, overwriting expected drafts, or storing unreviewed AI-generated content in a real business account.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
解决：
curl -F "media=@cover.jpg" "https://api.weixin.qq.com/cgi-bin/material/add_material?access_token=TOKEN&type=image"
cgi-bin/draft/batchget 拿当前 id 列表（可能为空，说明被删了），再决定是 update 已有还是重新 add：
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
解决：
curl -F "media=@cover.jpg" "https://api.weixin.qq.com/cgi-bin/material/add_material?access_token=TOKEN&type=image"
cgi-bin/draft/batchget 拿当前 id 列表（可能为空，说明被删了），再决定是 update 已有还是重新 add：
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
解决：
curl -F "media=@cover.jpg" "https://api.weixin.qq.com/cgi-bin/material/add_material?access_token=TOKEN&type=image"
cgi-bin/draft/batchget 拿当前 id 列表（可能为空，说明被删了），再决定是 update 已有还是重新 add：
The skill’s natural-language description, triggers, and operating instructions are all presented only in Chinese and are specifically framed around producing content in a微信公众号 style. There is no indication that users may choose another language or opt in to this locale-specific behavior.
Detected: suspicious.exposed_secret_literal