v0.3.1

Important Changed

(Add) AWorld CLI
(Add) AWorld Docs
(Add) Data Synthesis framework (beta)
(Add) Text2Training (Demo)

What's Changed

• Aworld CLI. @ahgpt #640 #642 #644 #670 #672 #691 #703 #715 #730
• AWorld docs by @rainsonGain in #667 #673 @wuman001 in #699 #700
• Data Synthesis. @rainsonGain in #686
• Text2Training. @rainsonGain in #686 #687 #688 #689 #690 #713
• Context improves. @ahgpt @tallate @SzekiHou #603 #608 #624 #637 #651 #657 #660
• Trajectory improves. @SzekiHou #649 #653 #658 #716
• Support PTC by @ahgpt in #628
• Sandbox improves. @kevinfeng030 #611 #612 #620 #625 #645 #712 #723
• Env client by @kevinfeng030 in #685 #705
• Storage: Qdrant Vector Search Support by @Anush008 in #647 by @mx-pai #737 by @NTLx in #738
• Models improve. @SzekiHou #665 #695
• Evaluation refactor. @rainsonGain #615
• Memory db by @xiudongxu in #599
• Runtime. #596 #652 #669 #680 #717 #727 #734 #720
• Log improves. #675 #718
• A2A client. @wangchunf in #589
• Train-related. @tallate in #588 @wangchunf #594 @rainsonGain #602
• Add HardGen and edit the wrong link of Aworld Tech Rport by @BingguangHao in #722
• Some fixes. #600 #610 #614 #616 #619 #621 #627 #634 #646 #679 #684 #704 #708 #710 #714 #731 #741

New Contributors

• @Anush008 made their first contribution in #647
• @wuman001 made their first contribution in #699
• @BingguangHao made their first contribution in #722
• @mx-pai made their first contribution in #737
• @NTLx made their first contribution in #738

Full Changelog: v0.3.0...v0.3.1

v0.3.0

Major Version Upgrade. The multi-agent foundation framework upgraded its abilities, and preliminary completion of the self-closed-loop on AWorld.

Important Changed

• (Add) The train module provided a unified API for agent training, running in code pattern.
• (Add) Supported agent to the agent server, such as Google A2A and Anthropic MCP, so endorsed streaming task, and provided new streaming task API
• (Add) Anthropic Skills ability is supported
• (Update) Context ability upgrade, supported context engineering

What's Changed

• AWorld training with VeRL, AReaL by @yuchengyue @yaoyueduzhen @rainsonGain in #517 #586 #590 @tallate in #528 @wangchunf in #543 #572
• AWorld agent server (experimental), supported A2A by @rainsonGain in #539 #570 @wangchunf in #576, supported to MCP @wangchunf in #560
• Skills ability support by @SzekiHou in #541 #585 @ahgpt in #542 @wuhulala in #573
• Context implement upgrade by @ahgpt in #525 #526 #532 @xiudongxu in #546 @wuhulala in #567 #571
• Streaming response API supported by @SzekiHou in #556
• Workspace support files up-download by @ahgpt in #561 #562 #564 #566
• Sandbox and MCP improve by @kevinfeng030 in #511 #536 #553 #581
• Trajectory improve by @SzekiHou in #550 #552 #577
• Trace improve by @wangchunf in #519 #533 asyncio monitor in #554 #582
• Agent improve by @rainsonGain in #551 @SzekiHou in #569
• Eventbus improve by @rainsonGain in #535
• Examples IMO trajectories and results by @JasonTse1989 in #514 #515 #516
• Examples add the visualwebarena example folder by @kevinhikali in #521
• Examples add Aworld tutorials of quick start by @rainsonGain in #559
• Some fix #512 #513 #518 #522 #524 #529 #530 #547 #574 #575 #584 #587

New Contributors

• @tomokinakamaru made their first contribution in #518
• @kevinhikali made their first contribution in #521
• @wuhulala made their first contribution in #567

Full Changelog: v0.2.8...v0.3.0

v0.2.8

Import Changes

Add dataset module, and remove replay_buffer module, unify trajectory to the Dataset (#494)
Add evaluations module implementation (#494)
Add a storage infrastructure module for other modules (#494)
Unify the logger and instrument in the AWorld project (#494)

What's Changed

• feature:
  (examples) Osworld agent submit by @ChenLLong in #472
  (EventRunner): cancel task when timeout by @SzekiHou in #474
  (ModelResponse): add reasoning content by @SzekiHou in #477
  (docs) update agent_training.md by @yuchengyue @rainsonGain in #489 add multi-actions parallel by @rainsonGain in #501
  (MCP) Adjust prefix by @kevinfeng030 in #493
  (Dataset): add dataset module and trajectory dataset by @SzekiHou in #494
  (Evaluation): add evaluations module by @wangchunf in #494
  (Storage): add AWorld storage module by @rainsonGain in #494
  (logger refactor): Unify logger in project by @rainsonGain in #494 add loguru instrument by @wangchunf in #504
  (AWorld) supported python3.10 by @rainsonGain in #498
  (Tool) return error message when tool_name not found in ToolFactory by @SzekiHou in #499
  (Agent):reset loop_step when terminate loop by @SzekiHou in #507
  (Trace) id gen by @wangchunf in #508

• Fix:
  Agent training by @yuchengyue in #475 by @rainsonGain in #481
  Memory fix by @tallate in #483
  Mcp tool fix by @xiudongxu in #496
  Human tool fix by @tallate in #482
  Stream model response by @SzekiHou in #485
  Env some fix: #487

Full Changelog: v0.2.7.1...v0.2.8

V0.2.7

Add

• Agent Training with 3rd train framework (#427)
• Environment (#427)
• HITL and its example (#419)
• IMO app (#336)
• Docs (#453)

Important Changed

• Context Update (#454 #340 etc.)
• Swarm supported nested(hierarchy) (#435) and standard workflow ability (#450)
• Memory add sqllite backend (#384) and other update (#402 #432)
• RuntimeEngine of multiprocess/ray/spark fix and enhancement for parallel and distributed running (#457)

What's Changed

• add imo example by @JasonTse1989 @eric68668 in #336 #338
• Environment by @eric68668 #427
• Train by @SzekiHou @rainsonGain in #427 #428
• Context by @tallate in #340 #359 @kevinfeng030 in #352 @ahgpt in #354 #358 #415 @SzekiHou in #396
• HITL and its example by @tallate in #419
• Trajectory by @SzekiHou in #381
• Docs by @rainsonGain in #453
• Swarm supported nested(hierarchy) and standard workflow ability by @rainsonGain #435 #450 @wangchunf in #449
• Memory add sqllite backend by @ahgpt in #384
• YAML-based config loader with in-file variables by @SNHuan in #376 #379
• Feature/gaia update by @JasonTse1989 in #369
• some refactor by @SyedaAnshrahGillani in #341 #350 @Mirza-Samad-Ahmed-Baig in #343 by @qingw-dev in #342 @rainsonGain in #364 #429
• some fix by @xiudongxu in #397 #406 @tallate in #437 @wangchunf #447 #422 #423

New Contributors

• @wwoodsTM made their first contribution in #339
• @SyedaAnshrahGillani made their first contribution in #341
• @Mirza-Samad-Ahmed-Baig made their first contribution in #343
• @Decem-Y made their first contribution in #360
• @JasonTse1989 made their first contribution in #369
• @SNHuan made their first contribution in #376
• @CzsGit made their first contribution in #393
• @JILLWONG made their first contribution in #444

Full Changelog: v0.2.6...v0.2.7

V0.2.6 (AWorld Architecture Upgrade)

What's Changed

• Event-driven supported by @rainsonGain in #94 #171 #227
• Sandbox supported by @kevinfeng030 in #106 #111 #112 #114 #116 #124 #130 #138
• Web server supported by @eric68668 in #125 #139 #156 #168 #175 #179 #195 #197 #203 #223 #281 #312 #316 #320 #295
• Swarm upgrade supported @rainsonGain in #196 #217
• Memory new feature supported @ahgpt in #163 #202 #222 #234 #264 #270 #285 #300
• Context supported by @tallate in #154 #169 #177 #193 #200
• Runtime state manager by @wangchunf in #166 #188 output_event_handler by @SzekiHou in #164 #174 callback by @kevinfeng030 in #194
• Ability upgrade of Gaia #102 #104 #105 #107 #115 #117 #119 #123 #133 #221
• MCP update supported by @kevinfeng030 #210 #245 #277
• Replay buffer by @SzekiHou in #319 #321
• Trace new feature by @wangchunf in #93 #120 #121 #127 #135 #136 #147 #160 #204 #296
• Feature/agent multi actions supported by @SzekiHou @wangchunf @rainsonGain in #293 #310
• Readme upgrade by @ZhuangCY in #280 #283
• AWorld examples change #324
• some fix: #146 #165 #167 #170 #172 #199 #248 #252 #318

New Contributors

• @tallate made their first contribution in #154
• @ultranity made their first contribution in #173
• @Raysin-cn made their first contribution in #238
• @mixiazhiyang made their first contribution in #88
• @ZhuangCY made their first contribution in #280
• @dependabot[bot] made their first contribution in #306
• @xiudongxu made their first contribution in #318

Full Changelog: v0.0.1...v0.2.6

v0.0.1

What's Changed

• Aworld framework and some issue fix by @rainsonGain in #1 #3 #5 #8 #10 #14 #23 #65
• app readme optimization by @yishanyoyo in #6
• gaia demo by @flashcp in #7 #9 #11
• Agent base and swarm by @rainsonGain in #17 #18 #22 #41
• mcp tool supported by @rainsonGain in #19
• mcp stdio and args by @kevinfeng030 in #21 #25 #32 #34 #37 #43 #45 #47
• Tracer, Metrics by @wangchunf in #20 #46 #50 #55 #77 #84
• llm model by @SzekiHou in #28 #29 #35 #44
• Aworld memory && Output by @ahgpt in #36
• Release gaia by @yuchengyue in #56 #57 #58 #60
• Dev trace UI by @wangchunf in #66 #67
• LLM model add ant provider by @SzekiHou in #69
• adding necessary docker mcp servers by @qingw-dev in #73
• Framework upgrade by @rainsonGain in #74
• Dev replay buffer by @wangchunf in #76 @SzekiHou in #78
• Gaia benchmark Docker env by @eric68668 in #79 #95

New Contributors

• @rainsonGain made their first contribution in #1
• @yishanyoyo made their first contribution in #6
• @flashcp made their first contribution in #7
• @kevinfeng030 made their first contribution in #12
• @ahgpt made their first contribution in #36
• @yuchengyue made their first contribution in #56
• @qingw-dev made their first contribution in #73
• @eric68668 made their first contribution in #79

Full Changelog: https://github.com/inclusionAI/AWorld/commits/v0.0.1