Introduction
Nexrall is an agentic coding platform — a set of agent-driven products that share one account, one wallet, and one model catalogue: generate full-stack apps from a prompt, write documents and decks with an AI writing agent, and run a coding agent inside your editor, terminal, and desktop.
Every Nexrall product is built around an agent loop — reason, call tools (read/write files, run commands, call APIs), observe the result, and repeat until the task is done. Each surface runs the loop that fits its job:
- Fullstack Agent Coding — describe an app in plain language, and an agent plans, writes, tests and deploys a real full-stack app (frontend, backend, database) with its own URL, then keeps iterating with you in chat.
- Nexrall Canvas — an AI writing agent that produces one polished document per session — reports, slide decks and dashboards with live charts — exportable to PDF.
- Mini Apps & Store — every app the agent builds runs on real infrastructure; keep it private or publish it for anyone to install in a tap.
For coding in your own workspace, the developer tools share one agent engine:
- Nexrall Code for VS Code — an embedded coding agent inside your editor.
- Nexrall CLI — the same agent in your terminal, for scripts and CI.
- Nexrall Work — the same agent as a terminal-free desktop app for working across folders and Office documents.
Why "agentic"?
Most "AI features" are one-shot: you ask, the model answers, and the context disappears. Nexrall is built the other way around — its agents are loops that plan, act, observe and correct, sharing a few primitives so long-running work doesn't lose context or repeat itself:
| Primitive | What it gives every agent |
|---|---|
| Memory | Conversations and a progress ledger so long-running agent work doesn't lose context or repeat itself. |
| Orchestration | Each agent runs an autonomous loop that plans, acts, retries and verifies its own work. |
| Tools | File edits, shell commands, HTTP calls, and connectors (Slack, GitHub, Google Sheets, and more) available to any agent that needs them. |
| Sub-agents | Delegation to specialised agents that run in parallel with their own tools and permissions — review, testing, research. |
| Oversight | Approval of high-risk actions and full visibility into what actually ran, not just what the agent claims. |
Where to start
- New to Nexrall? Start with the Quickstart.
- Want to understand how the pieces fit together? Read Core Concepts.
- Building something with the API? Jump to the API Reference.