Python SDK for AI agent monitoring, LLM cost tracking, benchmarking, and more. Integrates with most LLMs and agent frameworks including CrewAI, Agno, OpenAI Agents SDK, Langchain, Autogen, AG2, and CamelAI
Note: While this project may not have specific linting/formatting configurations, the repository uses shared development tools. Please see the root README.md for setup instructions if needed.
###Useful Commands:
Download Supabase CLI
npm i supabase --save-dev
or on Mac
brew install supabase/tap/supabase
login
supabase login
start supabase
supabase start
link with remote project
supabase link --project-ref <project-id>
pull changes from remote
supabase db pull
load supabase configuration:
supabase migration up
supabase db reset
supabase db reset --linked
deploy configuration to remote
supabase db push
create diff
supabase db diff
Setting up a new project
Setting up Auth
- Go to Authentication > URL Configuration
- Add
https://app.agentops.aias the Site URL - Add
https://app.agentops.ai/**into the Redirect URLs - Go to Authentication > Providers
- Enable GitHub (by creating a GitHub app and filling in the provided info)
- Go to Settings > Authentication > SMTP Settings
- Enable
Enable Custom SMTP - Fill in the required sender and SMTP Provider details
Setting up Webhooks
- Download Supabase CLI (See above)
- Run
supabase link --project-ref $Project-ID --password $db-password. The Project-ID can be found on supabase under settings > general > reference ID. If the password is not saved, it must be reset and stored in a secure location. - Create a .env file in the supabase folder
- Add .env variables:
ATTIO_API_KEY=
ATTIO_OBJECT_ID=
- Run
supabase secrets set --env-file ./supabase/.env - Run
supabase functions deploy - On the Supabase website, go to database > webhooks and click "enable webhooks"
- Create a new webhook that posts to the Supabase function. Add the authorization header from the dropdown. This should auto-populate the service role key
You should be set!