Security audit
Security checks for vulnerabilities and agentic risk
ClawScan marked this skill suspicious with medium confidence, mainly because local-photo workflows lack clear disclosure of external processing and the installation command uses an unpinned package; the scan was partial.
Before installing, consider whether you are comfortable with an unpinned installation command. Before using local or reference photos, check which services receive them and approve the transfer only if you are comfortable with their processing and storage. The scan was partial: it inspected all three listed components, but some local references could not be resolved, so the evidence is incomplete.
The installation command executes an unpinned npm package and retrieves skill content without an immutable revision. A compromised future release or repository revision could introduce code or agent instructions with access to the user's environment; this is a supply-chain weakness, although the command is not itself an MCP server reference.
The local-photo examples encourage an agent workflow that requires sending user images outside the local environment, without clearly explaining that transfer. Naming the services elsewhere provides some context, but does not establish informed approval to upload potentially sensitive photos.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
| 2048×2048, high | estimate first | 10.7¢ |
| Failed job | 0 | |
Check any request before you run it: `POST https://api.varg.ai/v2/estimate` with the same body, or
`estimate_cost` on the MCP. Follow the spend rules in the varg-ai skill (a cap, approved models,
estimate before every paid call).
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
-H "Content-Type: image/png" --data-binary @photo.png
# text to image (Flare)
curl -s -X POST https://api.varg.ai/v2/image -H "Authorization: Bearer $VARG_API_KEY" \
-H "Content-Type: application/json" -d '{
"model": "gpt_image_2_6",
"prompt": "…",
The instructions explicitly upload a local image to external storage and return a public_url, but provide no privacy checkpoint before the agent performs the transfer. Sensitive image content or metadata could leave the user's control, and the access restrictions and retention of the returned URL are unspecified.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
It returns `url`, `job_id` and `cost_credits`. An edit adds `"files": [{"url": "https://s3.varg.ai/…"}]`
to `input`. To use a local file, upload it first: `create_upload({"media_type": "image/png", "size":
<bytes>})`, then `curl -X PUT -H "Content-Type: application/octet-stream" --data-binary @photo.png
"https://api.varg.ai$proxy_url"` with the `proxy_url` it returned; the answer has the file's
`public_url`.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
# upload a local image -> {"url": "https://s3.varg.ai/...", ...}
curl -s -X POST https://api.varg.ai/v2/files -H "Authorization: Bearer $VARG_API_KEY" \
-H "Content-Type: image/png" --data-binary @photo.png
# text to image (Flare)
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
}'
# edit on Sunburst, custom size
curl -s -X POST https://api.varg.ai/v2/image -H "Authorization: Bearer $VARG_API_KEY" \
-H "Content-Type: application/json" -d '{
"model": "fal:openai/gpt-image-2.6/sunburst/edit",
"prompt": "…",
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
}'
curl -s https://api.varg.ai/v2/jobs/JOB_ID -H "Authorization: Bearer $VARG_API_KEY"
Model spec with every option: https://varg.ai/models/gpt_image_2_6.md. JSON schema (no key):
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
Model spec with every option: https://varg.ai/models/gpt_image_2_6.md. JSON schema (no key):
https://api.varg.ai/v2/tools/image?model=gpt_image_2_6.
## 7. Prompting text-heavy layouts
The recipes direct agents to submit user reference images through a remote editing workflow without explaining the privacy implications or requiring approval for external processing. The later requirement to obtain a depicted person's agreement is helpful, but does not explicitly address external storage, downstream processors, or retention.