Crypto Signal Checker

Turn a structured crypto signal into a clear, evidence-backed assessment.

• Live market context: check current BingX prices and the hourly trend.
• News context: recent coin and market headlines are provided to the model as unverified sources when the model is called.
• Portfolio context: optionally add your existing positions; no exchange connection or orders are needed.
• Liquidation context: prior four hours of Bybit liquidation totals when complete, fresh coverage is available.
• AI assessment: receive a structured Qwen decision with reasons and available evidence.
• Ready for workflows: use the Actor in Apify, via API, or with MCP.
• No OpenRouter key required: model usage is billed separately to the running Apify account.

Assess one structured USDT perpetual futures signal against recent BingX market data and Qwen. Get an approve, reject, or unknown result with the available evidence and a short reason. The Actor does not place orders, predict profits, or assess whether a trade is suitable for you.

What it checks

1. Validates the symbol, direction, prices, and signal geometry. LONG targets must be above the entry in ascending order, with the stop below it; SHORT targets must be below the entry in descending order, with the stop above it.
2. Retrieves closed hourly candles and recent trades from BingX USDT perpetual markets. It requires at least 50 consecutive hourly candles and a recent trade no more than 120 seconds old.
3. Rejects a signal if the current price is at or beyond its stop or first target. LONG requires a rising H1 close sequence and EMA20 above EMA50; SHORT requires a falling sequence and EMA20 below EMA50. An unclear trend returns unknown.
4. When those checks pass, sends the signal and market evidence to qwen/qwen3.8-max-0902 through OpenRouter for a conservative assessment. It checks a fresh trade again before returning the model's result.

approve is the model's assessment after the checks above, not a trade instruction or a guarantee. reject indicates a failed price or trend check, or a model rejection. unknown means the Actor could not make an assessment, for example because the input geometry is invalid, market data is missing or stale, or the model request or response failed. Invalid input can also fail the run before a dataset item is created. Model confidence is an uncalibrated judgment, not a probability of profit.

The Actor does not have channel history. News, user-provided positions and Bybit liquidations are included only when the model stage is reached; unavailable sources are explicitly marked. Liquidations cover Bybit only and are unavailable when the collector has gaps or stale coverage. Positions are not verified against an exchange, and news headlines are not verified facts. It checks the current price, but it does not know when the signal was published or whether a stop or target was reached earlier. It reports the current price distance from the planned entry; it does not verify execution, liquidity, or fees. It supports only BingX USDT perpetual markets. Market conditions can change immediately after the assessment.

Input

Provide a base symbol of 2–20 letters or digits, direction (LONG or SHORT), positive numeric entry and stop prices, and one to ten ordered numeric take_profits. Prices are in USDT. Optionally add original_text (up to 4,000 characters) for the model to compare against the structured fields. This comparison is model-based and is not independently verified. The Actor does not extract prices from free text.

Free and paid Apify accounts can run the Actor, subject to available credit and Apify access restrictions. Free-plan model access may be blocked by Apify anti-fraud protections. Apify currently lists a 10x OpenRouter markup for free accounts; check its current pricing. No separate OpenRouter account or API key is needed.

Optionally provide positions, an array of up to 50 objects such as {"symbol":"BTC","direction":"LONG"}. These entries are user-supplied and not verified; an omitted array means positions are unknown. Enter an empty array only if there are no open positions. Values are sent to the model and stored in the output. Never include account credentials or other sensitive details.

Example values below are hypothetical and are not a trade recommendation:

{
  "signal": {
    "symbol": "BTC",
    "direction": "LONG",
    "entry": 100000,
    "stop": 98000,
    "take_profits": [102000, 104000],
    "original_text": "BTC long, entry 100000, stop 98000, targets 102000 and 104000"
  }
}

Output

The default dataset contains one item when an assessment completes. Depending on the outcome, it includes status, reason, as_of, available sources, the normalized signal, market, and an assessment containing the model's decision, confidence, regime, risks, and short reasoning. An error code may explain an unknown result. Fields depend on how far the checks progressed.

Costs and data handling

The result fee applies per approve or reject; see the Pricing tab for the current amount. An unknown result has no assessment fee. Apify also charges the Actor start event ($0.00005 per GB, minimum one event), including for an unknown result or an ineligible run. Platform usage is included in the event pricing for standard runs. Model requests through Apify's OpenRouter proxy are charged separately to the running Apify account, including when the result is unknown; a signal rejected before that stage does not make a model request. Check the current prices and your account limits before running.

The Actor sends the symbol to BingX to fetch market data and, when the model stage is reached, searches Google News for recent headlines and fetches liquidation aggregates from the developer's read-only service. At that stage, the structured signal, optional original text and positions, market data, headlines, liquidation data, and evidence limitations are sent via Apify to OpenRouter. Apify stores run inputs according to its storage settings. The output contains the structured signal and supplied positions but does not copy the original text into the result dataset. Avoid putting credentials or sensitive information in signal fields. Protect your Apify credentials and review the providers' data policies.

API and MCP

You can run the Actor from Apify Console or via the Apify API: POST the input JSON to https://api.apify.com/v2/acts/genkenobi~crypto-signal-checker/run-sync-get-dataset-items with your Apify authorization. Keep tokens out of URLs, logs, and shared examples.

For AI agents, connect to the Apify MCP server, call genkenobi/crypto-signal-checker with the same input, and retrieve the dataset items from the resulting run. Use an authenticated client and treat its handling of input secrets as part of your security setup.

Example cost for 100 direct assessments

For 100 direct approve or reject results, multiply the current assessment price on the Pricing tab by 100, then add at least $0.005 for 100 Actor starts at the listed minimum of $0.00005 each. Model usage is billed separately whenever the model is called. An unknown result has no assessment fee, but may still incur start and model charges. Calling through the Crypto MCP Toolkit can add its own start fee. Check current prices, free-plan markups and your account limits.