crewai-econdash

CrewAI tools for EconDash. They call the econdash client, so they follow the same access rules. They do not open a free back door to the archive.

pip install crewai-econdash

Releases are tags crewai-vX.Y.Z on econdash-python (the X.Y.Z matches this package's version). A v* tag publishes the econdash client, not this package.

from econdash import EconDash
from crewai_econdash import EconDashCompareTool, EconDashSearchTool, EconDashSeriesTool

client = EconDash()                        # free trial, or ECONDASH_API_KEY from env
# client = EconDash(api_key="ed_...")
# from eth_account import Account
# client = EconDash(x402_signer=Account.from_key("0x..."), max_usd_per_call=0.05)

tools = [
    EconDashSearchTool(client=client),
    EconDashSeriesTool(client=client),
    EconDashCompareTool(client=client),
]

Each tool builds its own EconDash() when you pass no client.

What is free

Without a key and without a wallet the tools only receive the free trial: 10 countries (USA, CHN, DEU, JPN, GBR, FRA, IND, BRA, RUS, KOR), 50 indicators, 300 requests a day and 30 a minute per IP.

A country or indicator outside that trial makes the API answer 402. The tool turns that into a sentence for the agent and does not include the figures. Nothing is charged. A 429 becomes "Rate limit reached" with the wait time. Bad ISO3 codes and dates are answered with a hint, not an exception.

How to pay

• API key from an EconDash subscription (api_key= or ECONDASH_API_KEY): full archive, no per-call charge.
• Wallet on the client (x402_signer, tempo_account or solana_keypair) together with max_usd_per_call: pays per call, only when the price is within the cap. See the econdash README.

Example crew

examples/crewai_macro_research.py: a researcher agent collects GDP growth, inflation and unemployment for a country and compares it with USA and JPN; an analyst agent writes a short brief with every figure linked to its econdash.org chart.

export MODEL=gpt-4.1-mini OPENAI_API_KEY=...   # any model CrewAI supports
python examples/crewai_macro_research.py DEU