Originally published at nyria.io.
Every options alert service looks the same from the outside: a price, a screenshot, a channel you cannot see into. This guide is how to grade one from the inside, with its own posts, on a paper account that costs nothing, before a month's fee or a single live order. It ranks no services and names none.
What a usable alert contains
Five fields. An alert missing any of them is a trade idea, and a bot cannot place a trade idea.
Ticker The underlying, written as the exchange lists it: SPY, not "spiders". A call with no ticker is commentary.
Side and action Call or put, and whether the post opens or closes. BTO, STC and the rest, or plain words such as "out of SPY calls here". An exit that is implied, not posted, cannot be counted and cannot be traded.
Strike The exact strike, as listed. A strike that is not on the chain gets no contract, and a bot checks the chain rather than rounding for the caller.
Expiry Month and day. A call with no expiry takes the nearest listed expiration, which can be today's, even late in the day. If the caller leaves the date off, that is the trade they are posting.
Price The price the caller got. Without it there is nothing to compare a fill against, and the record is a list of tickers.
The notation does not matter; the fields do. Paste a week of the room's posts into the alert tester and see which ones read as an order and which get a refusal. Nothing is traded and no account is needed.
A worked example: one call, graded
An entry and its exit as a room would post them, and the three numbers a paper bot gives you for each.
Entry, posted at 9:47
BTO SPY 775c 11/20 @ 1.85
Exit, posted at 10:12
STC SPY 775c 11/20 @ 2.40
The paper bot reads the first as buy to open the SPY 775 call expiring Nov 20, sized from its own budget per entry, and the second as sell to close. Its Logs tab stamps the alert and the order, so for each call you hold three numbers side by side: the price the caller posted, the time the post arrived, and the price the paper bot filled at. Over a few weeks that table is the track record.
Grade it on paper, not on a screenshot
Run the service on a free paper bot for a few weeks and let the caller's real posts fire. Two ways in. If the caller publishes the calls as a Nyria strategy, subscribe and deploy a bot on Nyria Paper. If you run the server or the Telegram group, link it under Integrations, create a strategy with that channel as the alert source, and add an author filter so only the caller's posts count.
Read the paper numbers for what they are. Nyria Paper fills a market order at the quoted bid or ask, so it pays the spread but never slips, and a limit exit inside the spread fills at once at its own price. Paper P&L reads better than live. A service that cannot hold up on paper will not hold up live; one that does has passed a test, not every test.
Count the exits. A position closes only when a post reads as an exit, so calls the caller never closed in writing stay open in the paper account, and they would stay open at your broker.
Fill timing: a person versus a bot
The caller fills first, then types. A reader then sees the post, opens the broker and types again. That is two gaps, and a bot removes only the second one. Discord and Telegram stamp every message, so compare the caller's posted price with where the contract was trading at the stamp. A gap of seconds is normal. A gap of minutes, on a contract expiring today, is the whole edge.
A bot does not wait for a better price either. On Telegram, an entry Nyria handles more than a minute after it was posted is not placed at all; the strategy log says so, and a close in the same message still runs. Whatever the source, the paper fill tells you what the post was worth by the time anything could act on it.
Red flags
A promise about returns A percentage on a banner is a claim, not a record. The only record you can audit is the channel history: scroll it, count the entries, match each to its exit, and note the ones that never closed.
No exits posted A room that posts entries and leaves the exit to you is half a service, and the half you cannot grade. Partial exits and trims count if they are posted every time.
Alerts that change after the fact An edited or deleted post is a different record from the one that was traded. Nyria does not read an edit again, and editing or deleting a post after it trades does not undo the order, so what the bot logged is what was posted at the time.
A track record that is a screenshot A picture of a month's P&L shows one account on one day. It does not show which calls were skipped, which were closed late, or what a reader could have filled at.
Calls mixed into chat Alerts in a channel of their own, from one author, in one notation, can be read at a glance and by a bot. A channel where calls, chat and screenshots share the stream is slower to read by eye and harder for anything else to read at all.
Going live later, with your own broker
When the paper table holds up, deploy a bot on the live account, or edit the paper bot and pick the live account. The first live deploy asks you to accept the risk disclosure. Orders go to Charles Schwab, Tradier, tastytrade or Alpaca, in your own account, and the money never leaves your broker.
Paper did not test slippage, partial fills, broker rejections or assignment, so start with a budget per entry you can afford to lose in full, and hold capital for both the entry and the exit.
Questions
Can Nyria read a paid Discord I only belong to?
No. Nyria reads channels in servers you run, and groups or channels you administer on Telegram. If the caller publishes the calls as a Nyria strategy, you subscribe to it and deploy a bot in your own account.
How long should the paper test run?
Long enough to see the service in more than one kind of week. The docs ask for one to two weeks of paper before a first live bot; for grading a service you are about to pay for, a few weeks is a fairer sample.
What does the test cost?
Nothing. Paper is free. Live trading from $15 a month.
