Security audit
Security checks for vulnerabilities and agentic risk
The scanners rate this skill benign with medium confidence: its video-generation workflow appears coherent, though they flag extra installation and image-generation scope, an unpinned installer, and privacy concerns around cloud processing.
Before installing, consider whether you are comfortable sending character photos and dialogue to Varg for cloud processing, and review the extra clothing-image and broader plugin installation options. The installer version is not pinned. The verdict is benign, but confidence is medium and the scan is marked partial because two local references were unresolved.
Although image generation through Varg could support the declared workflow, this request specifically generates a clothing-only product photograph rather than a studio set, character reference, master shot, or derived dialogue shot. The supplied chunk therefore has a materially different purpose. Its use of Varg authentication is not itself a mismatch.
The supplied README shell commands perform installation and agent configuration rather than the declared podcast-video creation workflow. Installing the named skill could be a supporting setup step, but the alternative installation of an entire skill collection extends beyond that purpose. This assessment is limited to the supplied chunk; it does not establish what the installed skill subsequently does or imply malicious behavior.
The installation example executes the npm package skills without pinning its version, introducing a mutable dependency that runs with the user's permissions. A compromised future release could execute malicious code; however, this is an installation CLI supply-chain risk, not an unpinned MCP server as the finding states.
The README encourages supplying local character photos and dialogue but does not explicitly explain that these inputs leave the device for external processing. Listing the varg API and model providers gives some notice of cloud use, but users could still submit personal images or confidential scripts without understanding the disclosure.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
openclaw:
requires:
anyBins:
- curl
homepage: https://varg.ai
compatibility: >-
Needs a VARG_API_KEY with curl (see the varg-ai skill) or the varg MCP connection
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
Generate consistent-character podcast and dialogue videos with varg image generation + image-to-video with native audio.
Generation runs on varg: the `varg-ai` skill (REST at `https://api.varg.ai/v2`, sign-in and model details) or the varg MCP server (`https://mcp.varg.ai/mcp`: `generate_image`, `generate_video` or `call_tool`, then `get_job`). The REST calls below map one-to-one onto the MCP tools.
## When to use
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
## The workflow (8 steps)
Every generation returns a job. Poll `GET https://api.varg.ai/v2/jobs/{id}` until `status` is `completed`; the file is at `output.outputs[0].url`.
### Step 1: Write a short script
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
Generate the background/set as a standalone image with no characters. It becomes the base reference for every later shot.
curl -s -X POST https://api.varg.ai/v2/image \
-H "Authorization: Bearer $VARG_API_KEY" \
-H "Content-Type: application/json" \
-d '{
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
If a character needs clothing different from their reference photo, generate the outfit as a standalone image:
curl -s -X POST https://api.varg.ai/v2/image \
-H "Authorization: Bearer $VARG_API_KEY" \
-H "Content-Type: application/json" \
-d '{
The workflow directs the agent to upload character photos and submit dialogue to cloud generation without an explicit privacy or sensitive-data consent checkpoint. Cloud processing is disclosed, which reduces concealment concerns, but the later approvals concern stills and spending rather than disclosure of personal or confidential inputs.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
CRITICAL: Do NOT describe the characters' appearance in the prompt. Reference them by position ("left character from ref 1", "right character from ref 2 wearing outfit from ref 3"). Describing appearance conflicts with the ref image and produces a different person.
curl -s -X POST https://api.varg.ai/v2/image \
-H "Authorization: Bearer $VARG_API_KEY" \
-H "Content-Type: application/json" \
-d '{
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
# Close-up on character A
curl -s -X POST https://api.varg.ai/v2/image \
-H "Authorization: Bearer $VARG_API_KEY" \
-H "Content-Type: application/json" \
-d '{
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
CRITICAL: Put the dialogue line directly in the video prompt and ask for native audio. The model voices the line itself; no separate TTS.
curl -s -X POST https://api.varg.ai/v2/video \
-H "Authorization: Bearer $VARG_API_KEY" \
-H "Content-Type: application/json" \
-d '{