We think Crew AI is great. It's just that a lot of businesses we spoke to want the autonomy / decision making of an LLM but with guardrails you can dial up or down on both planning and execution. So we offer things like agent planning that evolves with your usage and declarative plan definitions. Our multi-agent runs also have persistent states and deterministic hooks so you can pause an agent, run more traditional code or solicit human input where needed.

Interesting - What's the biggest difference between this and Crew AI?

Can’t wait to kick the tires on this

#1 after 1 week on GitHub with perfect producthubt reviews? How much did all that cost you?

Shots fired hey :D We've been on Github since March actually and have a handful of design partners who do love the SDK so it wasn't hard to ask them to support our re-launch esp as the MCP feature release has been very useful for them. We delayed the announcement on here until now to try and polish the product a bit more first. Appreciate the spice though. Hope you're as gifted at finding bugs in our SDK hehe

That's super cool. Is data encrypted when using Portia in cloud mode?

For sure! If you want to know and in particular for sensitive data like oauth tokens please check this page out in our docs --> https://docs.portialabs.ai/security

Brilliant!

This is really interesting! Do you support Oauth with MCP tools as well?

Hi there (Portia dev here). Yes! we do support both Oauth and API auth types for MCP remote servers. Check out full guide here: https://docs.portialabs.ai/cloud-tool-registry