validate api script (#10217)

[ReynardoEW]

uses api-docs schemas, some error caused by wrong schemas definition so need to make changes there #10217

.env setting
BASE_API_URL=https://api.llama.fi
BETA_API_URL=
DISCORD_WEBHOOK_URL=
PRO_API_KEY=

[treeoflife2]

@ReynardoEW are we sending the json file in the discord? as discord message will be truncated when it's very big? everything else looks good.

[ReynardoEW]

@ReynardoEW are we sending the json file in the discord? as discord message will be truncated when it's very big? everything else looks good.

it send summaries not full json file something like this

pro api:
• total endpoints: 82
• pass: 77
• fail: 5
• success rate: 93.9%
• avg resp time: 1401ms

failed endpoints (5 of 5):
1. https://pro-api.llama.fi/api/emission/hyperliquid
2. https://bridges.llama.fi/bridgevolume/Ethereum?id=5
3. https://bridges.llama.fi/bridgedaystats/1667304000/Ethereum
4. https://bridges.llama.fi/transactions/1?starttimestamp=1667260800&endtimestamp=1667347200&sourcechain=Polygon&address=ethereum%3A0x69b4B4390Bd1f0aE84E090Fe8af7AbAd2d95Cc8E&limit=200
5. https://pro-api.llama.fi/coins/block/sample/sample

2025-07-21T06:19:06.323Z```

[treeoflife2]

@ReynardoEW it will only send the alert if there is failure right? imo json file would be better for debugging, but we can add it later if needed for jenkin jobs, or we can add link of file, as i believe filename will always same we can later use static link, but fine for now.

[ReynardoEW]

@ReynardoEW it will only send the alert if there is failure right? imo json file would be better for debugging, but we can add it later if needed for jenkin jobs, or we can add link of file, as i believe filename will always same we can later use static link, but fine for now.

@ReynardoEW it will only send the alert if there is failure right? imo json file would be better for debugging, but we can add it later if needed for jenkin jobs, or we can add link of file, as i believe filename will always same we can later use static link, but fine for now.

i have updated it to only send when error, yes i also think sending json link is better

[treeoflife2]

@ReynardoEW can we also automate the generation of this llms.txt file, it uses only pro-apis request/response information from the defillama-pro-openapi.json, everything else will stay the same
https://raw.githubusercontent.com/DefiLlama/api-docs/refs/heads/main/llms.txt
https://api-docs.defillama.com/llms.txt

[ReynardoEW]

@ReynardoEW can we also automate the generation of this llms.txt file, it uses only pro-apis request/response information from the defillama-pro-openapi.json, everything else will stay the same
https://raw.githubusercontent.com/DefiLlama/api-docs/refs/heads/main/llms.txt
https://api-docs.defillama.com/llms.txt

i think it should be possible but llms.txt file has to have description regarding the API stuff right? i think it's probably best to use stuff like https://github.com/mendableai/create-llmstxt-py and let it create it for us?

[g1nt0ki]

@ReynardoEW @treeoflife2 mostly good, add a minor fix here where beta test is skipped if url is not passed: d8c2e31