Welcome to Veyra
Veyra is an autonomous coding platform that plans, writes, executes and self-heals software changes. These docs cover everything from your first run to advanced multi-agent orchestration in production.
Quickstart
Run your first autonomous agent in under five minutes.
Agent system
How Veyra orchestrates 47 specialised sub-agents.
REST API
Trigger runs, fetch traces, manage workspaces.
Playground
Test prompts and models inside your dashboard.
Models & providers
Routing logic across Codex, GPT and Gemini.
Security
Sandboxing, secrets, audit and compliance posture.
What is Veyra?
Veyra combines deep code reasoning, deterministic execution sandboxes, and a hierarchy of specialised agents into one platform. Where traditional copilots suggest, Veyra decides, executes, verifies and recovers — autonomously.
- Autonomous loops — Veyra plans multi-step changes and runs them end-to-end.
- Self-healing — failed tests, type errors, or lint failures are recovered without human intervention.
- Provider-flexible — route between OpenAI, Google and Anthropic models per task.
- Audit-grade traces — every tool call, file diff and decision is recorded.
New to Veyra?
Start with the Quickstart, then read Core concepts to understand how runs, agents and providers fit together.