Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Social Media Automation
v1.0.0自动管理小红书和微信视频号多账号内容发布，支持内容模板、合规审查和数据追踪生成日报。
⭐ 0· 49·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The SKILL.md content matches the stated purpose (Xiaohongshu and 微信视频号 automation, templates, tracking, reports). However registry/_meta metadata mismatch (ownerId differs between listing and _meta.json) and the package declares no required credentials while the instructions and example code clearly require API tokens, cookies, and webhooks. That mismatch is unexpected and reduces trust.
Instruction Scope
Runtime instructions instruct the agent to read/write local credential files (.credentials/), load cookies, launch browsers for automation, call platform APIs, and send daily reports to Feishu. Those actions are within the skill's functional scope but the instructions also reference unspecified external endpoints and secret material without clear limits, and they instruct storing and loading keys/cookies from disk — which broadens the data the agent will touch.
Install Mechanism
No install spec and no code files in the package delivered to the agent (instruction-only). That minimizes installer risk since nothing arbitrary is downloaded or written by an installer. The SKILL.md includes example code, but no automatic install steps are declared.
Credentials
The skill declares no required env vars/primary credential, yet examples and scheduler use ACCESS_TOKEN, cookies, and a Feishu webhook (send_to_feishu). The doc also instructs storing secrets in .credentials and gives conflicting encryption guidance (recommend gpg/pass but example code uses cryptography.Fernet with a local key file). Requiring/manage many secrets without declaring them and suggesting a local encryption key file (.credentials/key.key) is disproportionate and insecure if not handled carefully.
Persistence & Privilege
The example includes a scheduler that runs periodic publishing and reporting jobs, which implies long-running processes if deployed. The skill does not request always:true and does not modify other skills, so it does not ask for elevated platform privileges — but the agent/operator must be aware this will perform recurring actions and will access stored credentials over time.
What to consider before installing
This skill appears to implement the promised social-media automation features, but several red flags mean you should be cautious: (1) The listing metadata ownerId differs from the _meta.json ownerId — verify the publisher/registry identity before trusting code. (2) The skill requires sensitive secrets (platform access tokens, cookies, Feishu webhook) even though it declares none; do not run it with real credentials until you confirm where/how secrets are stored. (3) The doc suggests storing an encryption key and encrypted credentials in .credentials/ and the example uses a local Fernet key file — storing the encryption key alongside encrypted data or in the repository is insecure. Prefer a dedicated secret manager or OS-level protected storage. (4) Browser automation can capture login sessions (including MFA cookies); review code that loads cookies and make sure it does not exfiltrate them. (5) There are contradictory recommendations (gpg/pass vs Fernet) — ask the author for a single, secure secret-handling plan and for explicit env var names for any webhooks/API tokens. Recommended next steps: verify the package owner, get the full source code (not only SKILL.md) and an explanation of secret storage and Feishu endpoint usage, and run the code in an isolated environment with least-privilege test credentials. If you cannot validate the author or confirm safe secret handling, do not install or run this skill with production credentials.Like a lobster shell, security has layers — review code before you run it.
latestvk976b0b5wfykmn3swktsj7txa183kr8j
License
MIT-0
Free to use, modify, and redistribute. No attribution required.