What would you like to build?
Describe your agent in plain English and I'll generate the flow.
⌘S save · ⌘⏎ run · ⌘B blocks · ⌘J chat · ⌘K code
100%
What should this agent do?
Describe your agent in plain English
Try:
or
Start from a template
Pick a pre-built flow or start from scratch
Deploy
Configure and deploy your agent to production.
Agent Configuration
Link deploy records to a saved agent so Chat Console uses the correct
runtime.
LLM Providers and Models
Deployment Method
Controls generated runtime endpoints:
/v1/chat/completions and/or
/chat.
Set a GitHub token in Settings to list repos and push worker code.
Settings
Configure your Akompani IDE connections and preferences.
Integrations
Connect to external services for deployment and version control.
GitHub
Push agent code to repositories
Not connected
repo scope. Create token
Cloudflare Workers
Configure deploy settings for generated packages
Optional Cloudflare D1 binding. See Workers Databases docs.
To bind Durable Objects, provide a script name for an
existing namespace. See Durable Objects docs.
LLM Providers
OpenAI-style multi-provider model routing with per-provider model stacks and API keys.
Pick an existing provider or create a new one.
Use an OpenAI-compatible chat completions path.
Session API Keys
Session-scoped keys used by static generation/chat/deploy tools.
Appearance
Data
Settings are stored in your browser. Calls go directly to provider APIs (GitHub/Cloudflare/LLM), not an Akompani backend server.