Skill flagged — suspicious patterns detected

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

ReelClaw

v1.0.0

Create, produce, and publish UGC-style short-form video reels at scale. Full pipeline: source UGC reaction hooks from DanSUGC, analyze app demos with Gemini...

0· 51·0 current·0 all-time
byDaniel Hangan@danielhangan
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The SKILL.md clearly requires a Gemini API key and integration with a DanSUGC MCP (including purchasing/posting/analytics), but the registry metadata lists no required env vars, bins, or config paths — a mismatch. Requiring a payments-capable MCP (dansugc) and a cloud AI key is coherent for the described pipeline, but the registry's empty requirements are inaccurate and misleading.
!
Instruction Scope
Runtime instructions perform network operations (upload videos to Google's Gemini File API, call DanSUGC MCP endpoints, upload final media to tmpfiles.org), instruct installing fonts by downloading/unzipping from a third-party site, and run package installs via apt/brew. The skill also enforces purchasing clips before download (billing flow) and includes raw proxy tooling (scrapecreators_raw) that can forward arbitrary requests — these expand the data surface beyond simple local editing.
!
Install Mechanism
There is no formal install spec (instruction-only), which reduces reviewability; however the instructions include direct downloads (curl unzip of TikTok Sans from cufonfonts.com) and call system package managers (brew/apt-get) if binaries are missing. Direct downloads from an unverified third‑party host and writing fonts to user font dirs are riskier than using vetted package/release hosts.
!
Credentials
SKILL.md requires GEMINI_API_KEY and a DanSUGC API key (dsk_ prefix) and uses those for video uploads, analytics, purchasing and posting — appropriate for the stated pipeline but not declared in the registry metadata. Requiring a payment-enabled DanSUGC key (and granting it the ability to create posts/purchase content) is sensitive: it can incur costs and publish content. The skill also uses external public hosting (tmpfiles.org) which may expose media publicly.
Persistence & Privilege
The skill is instruction-only, has no install spec and does not request always:true or modify other skills. It does suggest adding an MCP entry to the agent (DanSUGC) which stores an external API token for use, but that's normal for integrations and not an elevated platform privilege by itself.
What to consider before installing
This skill appears to implement the described reel-production pipeline, but there are multiple red flags you should check before installing or running it: - Clarify required credentials: the SKILL.md needs GEMINI_API_KEY and a DanSUGC API key/MCP (used for purchases, posting, analytics) but the registry metadata lists none. Do not provide keys until you confirm why each is required and what scope/permissions they need. - Billing/purchases: DanSUGC's purchase_videos tool deducts credits — ensure you understand pricing and that you trust the service before authorizing any purchase-capable key. - Data exfiltration/privacy: the skill uploads videos to Google's Gemini API and advises using tmpfiles.org to host public URLs. Uploaded media and derived analysis will leave your environment; avoid sending sensitive videos and review retention/privacy policies of those services. - Third‑party downloads: the font install fetches from a non-standard host (cufonfonts.com). Prefer official font sources or manual verification of downloaded files before running curl/unzip commands. - MCP raw proxy and scraping: the provided mcp__dansugc__scrapecreators_raw proxy can forward arbitrary scraping requests. Confirm intended use and limits to avoid unintended scraping or terms-of-service violations. - Least privilege & testing: if you try it, use scoped/test API keys, minimal balance, and a throwaway posting account. Run in a sandboxed environment and review all network calls and temporary files. If you need to proceed safely, ask the skill author for an updated manifest that accurately lists required env vars/binaries and for a list of exact endpoints the skill will call (and whether any keys are stored). If the author cannot provide that, treat the skill as higher risk and prefer manual execution of the provided commands after code review rather than granting an agent autonomous access.

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

latestvk97618z06540ma15jfhh099yxh83kybj

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments