Portrait Generator

Generate hyper-detailed AI portraits via ComfyDeploy Morfeo Portrait workflow. Control every facial feature: eyes, nose, lips, jawline, skin, hair, expressio...

This skill appears to be what it says (it POSTs parameters to ComfyDeploy), but there are two red flags: it implicitly requires an API key but does not declare that requirement, and the skill has no source/homepage or other provenance. Before installing: (1) ask the publisher to explicitly list required environment variables (e.g., COMFYDEPLOY_API_KEY) and to document how the key will be used; (2) verify the endpoint and deployment_id with a trusted ComfyDeploy account or documentation; (3) supply only a least-privilege API key or service account you control, and rotate it after testing; (4) avoid sending personally identifying images or real-person likenesses until you confirm privacy/consent and the destination's data retention policy; (5) request the full, untruncated SKILL.md to confirm there are no hidden steps that read local files or other env vars. If the publisher cannot provide provenance and a clear declaration of required credentials, treat the skill as untrusted.

Type: OpenClaw Skill Name: portrait-generator Version: 1.0.0 The skill bundle is designed for generating AI portraits via the ComfyDeploy API. All files and instructions align with this stated purpose, defining the API endpoint, request schema, and agent behavior rules. There is no evidence of data exfiltration beyond the intended API call, malicious execution, persistence mechanisms, or prompt injection attempts against the OpenClaw agent. Notably, the `SKILL.md` includes a security-conscious instruction: 'Never send fields outside the defined input schema', which helps prevent the agent from being tricked into sending arbitrary data to the external API.