Agent Framework v0.6.3 Released
Today’s the BIGGEST SMALL RELEASE we’ve ever made!
Agent Framework 0.6.3 is flying onto a computer near you, and like a good guest, it has not turned up empty handed:
Adds skill testing tools and documentation (see http://docs.fetch.ai/aea/skill-testing). See https://docs.fetch.ai/aea/skill/ for more about skills: they’re the key part of implementing the meat of an agent’s business logic for delivering economic value.
We’ve added human-readable log information for the ACN (Agent Communication Network) — the peer-to-peer networking that lets your agents talk to each other after discovery on the OEF. This hugely helps debugging ACN configuration issues and more.
Well requested, and well wanted, it’s the fantastic CLI upgrade command. With this, you can now auto-upgrade all vendor packages after releases making the journey forwards to the latest and greatest the easiest it has ever been.
We’ve extended the SOEF stuff to factor in the new chain identifiers that Wednesday’s OEF update. This supports our new Incentivised Testnet launching next week, and provides other networks for now and the future leading up to and beyond mainnet v2.
We’ve also added a CLI command to transfer wealth to make it super easy to move around testnet tokens
Added README.mds (or “readies”, as autocorrect insisted) to all agent projects so there’s a ton of guidance now on the registry about each of the projects that you can use and build from
On top of this, there are a bunch of tweaks, fixes and other bits, like additional validations on message flow to make finding bugs easier, especially when developing skills.
Next week, we’re launching 0.7.0 of the Agent Framework alongside the Incentivised testnet.
It’s a big time for autonomous economic agents, where even small is big. Get BUIDLing!
Release information here: https://github.com/fetchai/agents-aea/releases/tag/v0.6.3
You can get it on PyPI here: https://pypi.org/project/aea/0.6.3/
And the latest documentation is here: https://docs.fetch.ai/aea/
Want to know what an agent looks like and is made of?
Check out this link: https://docs.fetch.ai/aea/diagram/ for an overall architecture diagram and a description of all the bits and pieces that make an agent tick.
And with that in mind, then go here: https://docs.fetch.ai/aea/quickstart/)