Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Varg Ai
v2.0.5Generate AI videos, images, speech, and music using varg. Use when creating videos, animations, talking characters, slideshows, product showcases, social con...
⭐ 0· 162·1 current·1 all-time
byAlex@securityqq
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name/description (video/image/speech/music generation) match the declared requirement for VARG_API_KEY and the optional binaries (curl for cloud, bun/ffmpeg for local). The BYOK provider keys mentioned in references are optional and documented; requesting only VARG_API_KEY as primaryEnv is proportional to the stated gateway-based design.
Instruction Scope
SKILL.md contains explicit, service-related runtime instructions: check for $VARG_API_KEY, sign-in via email+OTP (curl calls to varg.ai), save credentials to ~/.varg/credentials and .env, and use curl to submit render jobs and poll results. These actions are within the scope of a rendering/gateway client. Two points to note: (1) the skill instructs the agent to ask users for their email and the OTP code (expected for the OTP flow) but explicitly warns against asking users to paste raw API keys. (2) It also recommends a network-based version check (curl raw GitHub) and an npx -y skills update — these perform network operations and may install packages remotely, so they should be treated as higher-risk operations and inspected before use.
Install Mechanism
There is no package/install spec (instruction-only), which is the lowest-risk pattern. However, the skill bundle includes setup scripts (scripts/setup.sh and scripts/setup.ts) and the SKILL.md recommends running 'bun scripts/setup.ts' or using 'npx -y skills update' for updates — both will execute code on the host. The skill does fetch a remote SKILL.md via curl for version checks. These are normal for CLI workflows but merit review of the scripts before executing them.
Credentials
Only VARG_API_KEY is required as a primary credential, which matches the gateway design. References document optional BYOK provider keys (FAL_KEY, ELEVENLABS_API_KEY, REPLICATE_API_TOKEN) for cost/billing control — optional and justified. The instructions do write the API key to ~/.varg/credentials and append to a .env file; that behavior is expected but you should be aware the key will be stored on disk (chmod 600 recommended by the skill).
Persistence & Privilege
always:false (no forced/global persistence). The skill writes its own credentials file (~/.varg/credentials) and suggests appending to project .env — this is normal for a CLI SDK. There is no instruction to modify other skills or system-wide agent settings.
Assessment
This skill appears to be what it claims: a varg.ai gateway client for creating media that expects a VARG_API_KEY and optionally bun/ffmpeg for local rendering. Before you install/use it:
- Do NOT paste secrets into chat. The skill will ask for your email and a 6-digit OTP during Option B (normal OTP sign-in), and it explicitly says not to ask users for raw API keys — follow that. If you already have a VARG_API_KEY, the skill asks you to export it in your shell (export VARG_API_KEY=...).
- The skill will save credentials to ~/.varg/credentials and append VARG_API_KEY to .env by default. If you prefer not to persist your key on disk, skip those steps or inspect the created files. Verify the contents of ~/.varg/credentials after setup (the SKILL.md save snippet uses a literal USER_EMAIL placeholder in one place — check that it stores your real email).
- The skill recommends running bundled scripts (scripts/setup.ts/sh) and using network operations (curl to GitHub and npx -y skills update). Treat those as operations that will execute code or install packages locally — inspect the scripts and consider running them manually in a controlled environment before allowing automated execution.
- Optional BYOK provider keys (Fal, ElevenLabs, Replicate, etc.) are documented and optional; only provide them if you understand billing implications (BYOK may avoid varg credits but charges the provider directly).
If you want extra assurance, open and review the setup scripts (scripts/setup.sh and scripts/setup.ts) before running them and confirm the version check/update commands point to trusted repositories. If anything in the scripts looks unexpected, do not run them.scripts/setup.ts:32
Environment variable access combined with network send.
scripts/setup.ts:13
File read combined with network send (possible exfiltration).
Patterns worth reviewing
These patterns may indicate risky behavior. Check the VirusTotal and OpenClaw results above for context-aware analysis before installing.Like a lobster shell, security has layers — review code before you run it.
latestvk973r0na5twx6f5efy6ky9380h83vq1z
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Any bincurl, bun
EnvVARG_API_KEY
Primary envVARG_API_KEY
