Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Wewe Rss Reader
v1.2.0微信公众号文章阅读技能。当用户发来公众号文章链接、需要阅读理解文章内容时使用。通过本地部署的 WeWe RSS 服务获取文章全文并转为纯文本。Requires: 无（调用本地 REST API，首次部署见 references/deployment.md）。
⭐ 0· 53·0 current·0 all-time
by@axelhu
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The skill is narrowly focused on fetching article text from a local WeWe RSS REST API at localhost:4001, which matches the description. However, the metadata declares no required binaries/env but the documentation expects tools (git, docker/docker-compose, python3) and a local containerized service, so there's a mismatch between claimed requirements and what deployment actually needs.
Instruction Scope
Runtime instructions tell the agent to call localhost endpoints (expected) but also to send the login confirmUrl/QR to a Feishu group via an unspecified 'message' tool and to run docker compose in ~/.openclaw/workspace/wewe-rss/. The skill directs the agent to perform network activity to an external chat endpoint (sending a QR URL/image) and to run commands in a specific user path — both are actions outside simple read-only article fetching and should be considered sensitive.
Install Mechanism
There is no install spec (instruction-only), which keeps risk low. But references/deployment.md instructs cloning a GitHub repo, running docker compose, and launching python3 api_server.py. Those practical installation steps are real requirements but are not declared in the skill metadata (no required binaries listed).
Credentials
The skill declares no required environment variables or credentials, yet the docs mention AUTH_CODE and the need to send the confirmUrl to Feishu and to have a logged-in WeRead account. The skill does not document or require Feishu credentials, WeRead credentials, or other integration tokens — this omission creates ambiguity about what secrets or integrations the agent will need or attempt to use at runtime.
Persistence & Privilege
The skill is not always-enabled and does not request persistent platform privileges. It tells the agent how to start containers and a local Python process, which will create persistent services on the host, but that is consistent with a local helper service and not an intrinsic privilege escalation within the skill bundle itself.
What to consider before installing
Key considerations before installing/using this skill:
- Trust and isolation: The skill expects you to run a locally-hosted service (wewe-rss on :4000 and a REST adapter on :4001). Only deploy the referenced repository (https://github.com/AxelHu/wewe-rss) in an environment you control; review api_server.py and docker-compose.yml before running them.
- Missing declarations: The skill metadata doesn't list required binaries, but deployment requires git, docker/docker-compose, and Python 3.8+. Ensure these are available and run the service in an isolated environment (VM/container) if you have concerns.
- QR/URL sharing: The login flow returns a confirmUrl/QR that the SKILL.md instructs to send to a Feishu group. Sending this URL/image to an external chat will expose an authentication flow (and possibly allow the holder to complete login). Only send it to recipients you trust, and ensure the agent uses an authenticated Feishu integration you control.
- Workspace path and commands: The docs suggest running docker compose in ~/.openclaw/workspace/wewe-rss/. Make sure you understand and control that path; the agent's instructions to run shell commands could modify your local environment.
- What to verify: (1) inspect api_server.py and docker-compose.yml for any unexpected outbound endpoints or credentials; (2) confirm no hidden remote endpoints are used (the skill should only call localhost:4001/4000); (3) decide how the agent will send the QR (provide a controlled message tool or do that step manually).
Given these mismatches between documentation and declared requirements, proceed with caution — review the deployment artifacts and prefer running the service in an isolated environment before giving the agent runtime access to control it.Like a lobster shell, security has layers — review code before you run it.
latestvk977yh13r1fd8jqyyaqt591mm5840n6t
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
📖 Clawdis