AI agents need data that's already shaped for decisions—not five API keys and a parsing layer. I built crypto-mcp-toolkit, an Apify Actor that exposes crypto market data and backtested Telegram signal-channel statistics through one pay-per-call interface.
The 40-second demo is here:
https://x.com/DanteThomason/status/2103487147085074728
It exposes 12 tools, costs $0.005 per successful tool call, and does not require a separate crypto API key.
The Actor accepts one tool call per run:
curl -X POST "https://api.apify.com/v2/acts/genKenobi~crypto-mcp-toolkit/run-sync-get-dataset-items?token=YOUR_TOKEN" \
-H "Content-Type: application/json" \
-d '{"tool":"get_crypto_price","arguments":{"coin":"bitcoin","currency":"usd"}}'
The result is JSON that an agent can consume directly.
Find a channel first, then request its statistics:
{"tool":"channel_find","arguments":{"name":"@some_signals"}}
Then pass the returned channel ID to get_channel_stats. For a quick overview, use channel_leaderboard with a limit.
The important part is the evaluation method: expectancy is reported with a one-sided 95% confidence bound, not just a headline win rate. In my current sample, the median win rate is 76%, while median expectancy is negative.
Use the Apify API from n8n, Make, LangChain, or the Apify MCP server. One tool call in, one structured JSON result out.
I built and maintain this Actor. It is available here: https://apify.com/genkenobi/crypto-mcp-toolkit
Feedback is welcome—especially from people using crypto data in agents or checking paid signal groups before subscribing.