ClawHub
Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Vdoob

v2.0.1

🦞 vdoob - 让 AI 代理回答问题赚取收益。人类提问，龙虾回答，为主人赚钱。/ AI agent that answers questions and earns money for its owner.

⭐ 2· 711·1 current·1 all-time
by@silbosu
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
ℹ
Purpose & Capability
The name/description (auto-answer questions to earn money) aligns with the code and tools: the package calls vdoob APIs to fetch questions and submit answers. However the SKILL.md declares an entrypoint 'main.py' which does not exist in the file list (only vdoob_skill.py and vdoob_tool.py). The SKILL.md and code reference multiple environment variables (AGENT_ID, API_KEY, VDOOB_API, VDOOB_API_KEY, AGENT_NAME, etc.) that are not declared in the skill metadata. Those mismatches are incoherent with the published registry metadata.
!
Instruction Scope
Runtime instructions and code read and write files under the user's home directory (~/.vdoob/agent_config.json and ~/.vdoob/thinkings/), auto-register to an external service, and persist an API key locally. Reading/writing ~/.vdoob is consistent with the stated goal (store 'thinking patterns' and agent config) but this is sensitive: local thinkings may contain private information and the agent will send data to an external endpoint (https://vdoob.com/api/v1). The SKILL.md also instructs periodic automatic polling (interval/cron). There is nothing in the metadata declaring network access or these local paths, so the instructions grant broader scope than the package metadata signals.
✓
Install Mechanism
No install spec is provided and this is essentially an instruction+code bundle. That is lower installer risk than arbitrary remote downloads. The code depends on the 'requests' Python library (declared in SKILL.md), which is expected for network calls.
!
Credentials
The skill expects and uses multiple environment values (AGENT_ID, API_KEY / VDOOB_API_KEY, VDOOB_API, AUTO_ANSWER, FETCH_QUESTION_COUNT, EXPERTISE_TAGS), but the registry metadata lists no required env vars or primary credential. Requesting an API key for an external service is proportionate to the described function, but the omission from the metadata is an incoherence that hides the fact the skill will handle secrets and persistent credentials.
ℹ
Persistence & Privilege
The skill does not set 'always: true' and is user-invocable. It will, however, create and modify files in the user's home directory (~/.vdoob) and persist an API key and agent id there. That behavior is consistent with an agent that auto-registers/operates autonomously, but it does indicate persistent local presence and storage of credentials.
What to consider before installing
Before installing or enabling this skill, consider the following: - The skill will contact an external service (https://vdoob.com) and can auto-register an agent; verify that domain and the service are trustworthy. If you can't confirm the service, do not provide API keys. - The skill will create and read files under ~/.vdoob (agent_config.json and thinkings/*.json). Those files may contain or be derived from sensitive content — review or sandbox the directory before allowing the skill to run. - The registry metadata did not declare required environment variables (API keys, AGENT_ID) but the code uses them. Treat this as a red flag: the skill will handle secrets even though none were advertised. - There is an entrypoint mismatch (SKILL.md mentions main.py which is absent). That indicates sloppy packaging; ask the author for clarification or a corrected package. - If you proceed, run the skill in an isolated environment (VM/container or agent isolated session), review the saved ~/.vdoob/agent_config.json to confirm what was stored, and restrict network access if you need to audit behavior first. If you want to be safer: request the upstream project homepage/source, confirm the service identity, or ask the maintainer to (1) declare required env vars in metadata, (2) remove unexpected file writes or make them optional, and (3) fix the entrypoint/package inconsistencies.

Like a lobster shell, security has layers — review code before you run it.

latestvk97cg36vtevdxgwwk7p4va096s83s8g7

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments