
from suna19,777
Run, compare and manage AI model inferences via the Replicate API for image, video, and audio generation and other ML inference tasks.
This skill provides structured guidance and best-practice workflows for using the Replicate API to run ML models. It helps an agent choose appropriate image/video/audio models, fetch model schemas, create predictions, poll results, and download outputs safely. Practical recommendations cover seeding, concurrent predictions, webhook usage, and handling expiring output URLs.
Use this skill when a user requests generation or transformation of media (text-to-image, image editing, upscaling, background removal), text or audio synthesis, or any external ML inference where Replicate hosts a suitable model. Also trigger when asked to compare model options or automate repeated model runs.
Likely compatible with agents that can make HTTP requests and manage secrets (generic HTTP-capable agents like Claude Code, Copilot-style assistants, or custom agent runtimes).
The suna-replicate skill had a null/empty SKILL.md body and no bundled scripts. The GitHub source URL returns 404, suggesting the skill was either removed or the path changed. Without any content to evaluate, the skill provides no instructions, no structure, and no utility. It appears to be a dead link.
No content to audit. The source repository path no longer exists (404). This skill should be considered for removal or re-discovery.