TradingView alerts and AI agents send the signal. Vorda checks it against your rules and executes on your broker or exchange, with full transparency.
Trading somewhere else? Tell us which broker or exchange you need. The router doesn't care where the order ends up. That's the point.
Your TradingView strategy or indicator sends its webhook to your Vorda endpoint. Same alert, same JSON you already use. Nothing to rewrite.
IN action · symbol · qtyIs the bot known? Is the symbol real? Is the size sane? Did this exact alert already arrive? Bad requests stop here, and the log says why.
CHECK schema · auth · dedupePosition caps, symbol allowlists, account limits, market hours. You set them once per bot. Vorda enforces them before anything reaches your broker or exchange.
GATE size · limits · sessionVorda sends the order to your broker or exchange and records the answer. Filled, rejected, or blocked, you see the reason. Not a mystery.
OUT order → broker/exchangeConnect Vorda as an MCP server and your agent can check positions, place orders, and read execution logs, all inside the same risk rails as every TradingView alert. The agent proposes. Vorda validates, executes, and writes it all down.
Most automation tools fail quietly. You find out when the position is wrong. Vorda works the other way: it writes down every decision it makes, and every control stays in your hands.
Test the full flow before any capital is live.
Go live on one account, cleanly.
Scale across accounts and strategies.
Desk-grade capacity and onboarding.
The most common causes are malformed JSON, a wrong symbol, a size cap, a duplicate alert, a closed market, or a broker or exchange rejection. In Vorda, every blocked order carries its reason in the execution log, so you can see which gate stopped it instead of guessing.
No. You point your existing alert at a Vorda webhook URL and keep your JSON payload as it is. Validation, routing, and broker or exchange mapping all happen on Vorda's side.
No. API keys are scoped to trading only, and Vorda rejects keys with withdrawal permissions during setup. Credentials are stored encrypted and used for order routing, nothing else.
The webhook fails loudly, not silently. TradingView records the delivery failure, and no order can appear out of nowhere because orders only ever come from a received, validated payload. When service resumes, nothing replays unless your duplicate rules allow it.
One click pauses the bot. One click pauses everything. While paused, the router rejects incoming alerts for that scope and logs each rejection, so no order leaves Vorda.
Request it, then test everything else in the sandbox while you wait. Broker and exchange support grows without changing your alert flow, so when yours ships, you connect it and go.
Yes. Connect Vorda as an MCP server and the agent can check positions, propose orders, and read execution logs through scoped tools. Every agent order passes the same validation as a TradingView alert, and the agent never sees your API keys.
MCP is the open protocol AI agents use to call external tools. You only need it if you want ChatGPT, Claude, or another agent to act on your account. TradingView webhook automation works without it, and both can run side by side.
Free. No card. Paper fills until you flip the switch.