Skill flagged — suspicious patterns detected

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

Byted Las Video Edit

v1.0.0

Extract video clips from long videos based on natural-language descriptions. Use this skill when user needs to: - Extract highlights or specific scenes from...

0· 87·0 current·0 all-time
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
Purpose & Capability
The skill's name/description match the implementation: scripts/skill.py submits/polls tasks to operator.las.<region>.volces.com, accepts video_url, reference images, and output TOS paths, and returns clip metadata. These requirements are coherent for a video-editing integration. However, the registry metadata lists no required environment variables while both SKILL.md and the script require LAS_API_KEY (and optionally LAS_REGION / LAS_API_BASE). That metadata omission is an inconsistency the user should resolve before trusting the package metadata.
Instruction Scope
SKILL.md and scripts/skill.py keep to the stated purpose: they assemble JSON payloads, validate URLs (http/https/tos), avoid private-IP targets, read an API key from the environment or a local env.sh file in the current working directory, and call the documented submit/poll endpoints. The script does DNS resolution (socket.gethostbyname) to block private IPs and validates user-supplied URLs. It may read user-specified JSON files and env.sh in the CWD, but it does not appear to read unrelated system files or attempt to exfiltrate local files beyond values the user supplies.
Install Mechanism
No install spec — instruction+script only. No downloads, installers, or archive extraction are present, so nothing is written to disk aside from whatever the user runs (and any output files the script writes by design).
!
Credentials
The script demands a LAS_API_KEY (and optionally LAS_REGION / LAS_API_BASE), which is appropriate for authenticating with the LAS operator service — a single service credential is proportionate. The concern is that the registry's declared requirements list zero environment variables while SKILL.md and the code clearly require LAS_API_KEY (and mention LAS_REGION/LAS_API_BASE). Also the script supports reading env.sh from the current working directory; if users keep other secrets in that file they could be read unintentionally. The omission in the registry metadata reduces transparency about what secrets will be required.
Persistence & Privilege
The skill does not request always: true, does not modify other skills, and does not claim system-wide persistence. It's user-invocable and allows autonomous invocation by the agent (default), which is normal. There is no indication it writes config beyond its own outputs.
What to consider before installing
Key points to consider before installing/using: - Metadata mismatch: The registry claims no required env vars but SKILL.md and scripts/skill.py require LAS_API_KEY (plus optional LAS_REGION / LAS_API_BASE). Ask the provider to update registry metadata so you know what secrets you'll need to supply. - Credential safety: Only provide a LAS_API_KEY that is scoped/minimal for this service; avoid reusing long-lived or broadly privileged keys. Prefer creating a dedicated key for this skill and be prepared to revoke it if needed. - env.sh and working directory: The script will try to read env.sh in the current working directory if LAS_API_KEY is not in the environment. Do not place unrelated secrets in env.sh. Run the skill from a directory that contains only the intended env.sh or use environment variables instead. - Network behavior: The script performs DNS resolution to block private IPs and then sends URLs (and the API key in Authorization header) to the documented operator endpoints. Confirm those endpoints (operator.las.cn-beijing.volces.com / operator.las.cn-shanghai.volces.com or your explicit LAS_API_BASE) are expected. Because the operator may fetch referenced URLs (e.g., reference images or tos:// paths), avoid passing URLs that expose internal resources you don't want the operator to access. - Provenance & trust: Source/owner is unknown. If you need higher assurance, request the upstream source repository, a signed release, or a vendor statement. Consider running the script in an isolated environment (container or VM) and testing with a non-production API key first. - Ask for small fixes: Request that the package registry metadata be corrected to declare LAS_API_KEY and optional env vars, and ask the maintainer to document the minimum required key permissions. If you can't obtain these assurances, treat installation as higher risk.

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

latestvk979txexje4pv7wzt2kr6g3y2d83kwmf

License

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

Comments