Problem Even if you are a PRO user using the API_KEY, you are rate limited as if you don't use API_KEY.

Source of the Problem and Solution This looks to me like a bug in the code. The api_dict is not used anywhere as part of headers when elaborating/creating the requests. The get_response(...) function is expected aheaders as part of its arguments but it's not passed when called in any of the API endpoint wrappers/functions. I tested this in the get_metric_timeseries function. I did the following change in my local:

Before:
response = self.get_response(url, params=payload)

After
response = self.get_response(url, params=payload, headers=self.api_dict)

and it worked beautifully. My requests are rate-limited now based on the API KEY. I guess a similar thing needs to be done to the rest of the functions and that should fix the problem.

Summary

New /chains Endpoint

Adding another method to the Defillama class to consume the output from the /chains API endpoint.

This will allow us to programmatically get the TVL and/or names of all protocols for a specific chain.

Sorting the results to ensure that they're returned consistently between runs.

Empty Check for get_chain_tvl_timeseries

When grabbing data from get_chain_tvl_timeseries() for certain chains, the global TVL data can come back empty (ran into this for OntologyEVM)

To fix this, check to see if the global TVL results are empty and return an empty pd.DataFrame before assigning column names.

Tests

New endpoint Results before the check

Results after the check is added

get_chain_tvl_timeseries() works for all chains returned when calling the /chains endpoint on Defillama