Live, current data for your CrewAI agents and crews.
Point a CrewAI agent at one MCP URL - or POST one REST batch - and the whole crew gets 87 keyless tools of fresh, current data, each datapoint signed and stamped with its source and observed-at time.
The call
No SDK needed to try it - hit the keyless REST batch endpoint from a terminal and feed the same JSON into a CrewAI tool. Each result carries a source, observed-at time, and an Ed25519 signature.
curl -s -X POST https://dynamicfeed.ai/v1/batch \
-H 'Content-Type: application/json' \
-d '{"calls":[
{"tool":"weather_forecast","args":{"city":"Sydney"}},
{"tool":"check_vulnerability","args":{"package":"lodash","version":"4.17.10","ecosystem":"npm"}}
]}'Sample response
{
"mcpServers": {
"dynamic-feed": {
"url": "https://dynamicfeed.ai/mcp"
}
}
}Why live data
CrewAI agents reason over whatever context you hand them, and a base model's training is frozen at a cutoff - so a crew asked about today's weather, a fresh CVE, or a current LLM price is guessing. Wiring in a live-data layer turns that guess into a citeable fact. Connect the keyless MCP endpoint (Streamable HTTP) as a tool source for your crew, or call the keyless REST batch endpoint from inside a custom tool - either way there is no API key to provision and no per-agent setup. Every datapoint comes back with its source, an observed-at timestamp, a freshness marker, and an Ed25519 signature. Be precise about what that signature means: it is proof-of-existence and tamper-evidence - proof that we reported value X at time T and that the bytes were not altered in transit. It is not a claim that X is true, accurate, or safe to act on. Your crew still owns the judgement; we make the inputs verifiable and auditable.
Use it for
- Give a CrewAI research crew live weather, hazards, and current events instead of stale training data
- Have a security crew check a package against fresh CVE and known-exploited-vulnerability data before approving a dependency
- Ground a crew's final answer in citeable facts with a source and observed-at timestamp on each datapoint
- Add one shared MCP tool source so every agent in the crew calls the same 87 keyless tools
FAQ
Do I need an API key to connect CrewAI?
No. The MCP endpoint at https://dynamicfeed.ai/mcp and the REST batch endpoint are both keyless and free for fair use. Add the URL as a tool source for your crew and you are done.
How do I wire it into a crew - MCP or REST?
Either. Add the MCP URL as a Streamable HTTP tool source so every agent shares the same 87 tools, or call POST https://dynamicfeed.ai/v1/batch from inside a custom CrewAI tool when you want one self-contained function. The curl above is the exact REST shape.
What does the signature on each datapoint prove?
It is proof-of-existence and tamper-evidence: proof that we reported a given value at a given time and that the bytes were not altered. It is not a claim that the value is true, accurate, or safe to act on - your crew still makes the call.