Auto-GPT v0.4.7 introduces initial REST API support, powered by e2b's agent protocol SDK. It also includes improvements to prompt generation and support for our new benchmarking tool, Auto-GPT-Benchmarks.
We've also moved our documentation to Material Theme at https://docs.agpt.co. And, as usual, we've squashed a few bugs and made under-the-hood improvements.
What's Changed
- Integrate Auto-GPT with Auto-GPT-Benchmarks by @merwanehamadi in #4987
- Add API via agent-protocol SDK by @ValentaTomas in #5044
- Fix workspace crashing by @merwanehamadi in #5041
- Fix runtime error in the API by @ValentaTomas in #5047
- Change workspace location by @merwanehamadi in #5048
- Remove delete file by @merwanehamadi in #5050
- Sync release v0.4.6 with patches back into
masterby @Pwuts in #5065 - Improve prompting and prompt generation infrastructure by @Pwuts in #5076
- Remove append to file by @merwanehamadi in #5051
- Add categories to command registry by @Pwuts in #5063
- Do not load disabled commands (faster exec & benchmark runs) by @lc0rp in #5078
- Verify model compatibility if OPENAI_FUNCTIONS is set by @Pwuts in #5075
- fix: Nonetype error from command_name.startswith() by @lc0rp in #5079
- slips of the pen (bloopers) in autogpt/core part of the repo by @cyrus-hawk in #5045
- Add information on how to improve Auto-GPT with agbenchmark by @merwanehamadi in #5056
- Use modern material theme for docs by @lc0rp in #5035
- Move more app files to app package by @collijk in #5036
- Pass TestSearch benchmark consistently (Add browse_website TOKENS_TO_TRIGGER_SUMMARY) by @lc0rp in #5092
- Bulleting update & version bump by @lc0rp in #5112
New Contributors
- @ValentaTomas made their first contribution in #5044
Full Changelog: v0.4.6...v0.4.7