Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Hooked
v1.0.0Create AI-powered videos via the Hooked Video API. Script-to-video, prompt-to-video, UGC ads, TikTok slideshows, avatar selection, voice cloning, and trend d...
⭐ 0· 79·0 current·0 all-time
by@ycfra
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
Name/description match the SKILL.md and README: this is a Hooked video API integration. However the registry metadata claims no required environment variables or primary credential while the SKILL.md and README explicitly instruct the user to set HOOKED_API_KEY and use an x-api-key header. The package.json/README list homepages and repo URLs (hooked.so, GitHub) but the skill metadata shows source/homepage as unknown/none — an inconsistency that warrants verification of the publisher before installing.
Instruction Scope
SKILL.md gives concrete HTTP endpoints and workflow instructions (create videos, list avatars/voices, check status). It also describes optional webhooks (agent will POST video.completed including downloadUrl) and a productUrl parameter that the service may 'extract visuals' from. These behaviors are reasonable for a video API but introduce data-exfiltration risk if webhooks point to untrusted endpoints or if you pass sensitive/private URLs as productUrl. The instructions also require an environment variable (HOOKED_API_KEY) not declared in the registry manifest — the runtime agent will need that secret to operate.
Install Mechanism
There is no install spec in the registry (instruction-only), which is low-risk. README suggests npx/clawhub install methods and package.json exists, but no code files are included in the package provided. This mismatch (packaging metadata present but no install artifact in the registry entry) is a packaging/metadata inconsistency to check, but not directly malicious.
Credentials
SKILL.md requires HOOKED_API_KEY (x-api-key header) yet the registry metadata lists no required env vars or primary credential. Asking for a single API key for the external service is proportionate for this purpose, but the failure to declare it in the registry is a red flag (transparency/least-privilege lapse). Additionally, the use of webhooks and productUrl implies the skill will transmit or cause the service to fetch externally-hosted content — ensure you won't expose private URLs or data.
Persistence & Privilege
The skill is not always-enabled and uses default agent invocation settings. It does not request elevated or persistent platform privileges in the registry metadata, and there is no install script in the registry entry that would modify other skills or global config.
What to consider before installing
Before installing: verify the skill's publisher and source (confirm hooked.so and the GitHub repo listed in package.json actually host this skill). Expect to need an API key (HOOKED_API_KEY) — store it securely and do not paste it into chat. Only provide webhook URLs you control (they will receive download URLs and event payloads). Avoid passing private/internal URLs as productUrl because the service may fetch or include their content. Because the registry metadata omits the declared API key, treat this as a transparency/packaging issue: prefer skills whose manifest explicitly lists required secrets. If you decide to install, rotate the API key afterwards and monitor for unexpected activity; if unsure, contact Hooked support or use their official docs to confirm endpoints and behavior.Like a lobster shell, security has layers — review code before you run it.
latestvk97bsfg5f74mwvj6y6evq1z58983a9yy
License
MIT-0
Free to use, modify, and redistribute. No attribution required.