Search and run hundreds of AI apps — image generation, video, TTS, web search, LLMs
Use the belt CLI — run hundreds of AI apps, manage knowledge, search skills, connect MCP servers. Purpose-built CLI interface for agent workflows — typed inputs, schema validation, no raw API calls needed.
Build and deploy a custom AI agent on belt — define purpose, model, tools, system prompt, and deploy as a reusable agent. Use when the user says 'agentify', 'make an agent', 'create an agent', 'build a bot', or wants to turn a workflow into a deployed agent…
Build and deploy an inference.sh app from a conversation — scaffold, implement, test, deploy, and configure pricing. Use when the user says 'appify', 'make an app', 'deploy this as an app', or when a working API integration, model wrapper, or processing…
Build and deploy an inference.sh flow — chain multiple apps into a pipeline with wired inputs/outputs. Use when the user says 'flowify', 'make a flow', 'chain these apps', 'build a pipeline', or when multiple apps should run in sequence with outputs feeding…
Turn a working solution into a permanent, tested skill — extract pattern, interrogate for depth, create SKILL.md, publish to registry. Use when the user says 'skillify', 'make this a skill', 'remember this as a skill', or when a multi-step workflow was…
Run a simulated expert panel on a strategic question — brief 4-8 named experts in parallel, synthesize into consensus table
Save and recall knowledge — discoveries, insights, references, preferences