hud-python 0.3.0__py3-none-any.whl → 0.3.2__py3-none-any.whl

{hud_python-0.3.0.dist-info → hud_python-0.3.2.dist-info}/METADATA

@@ -1,9 +1,9 @@
 Metadata-Version: 2.4
 Name: hud-python
-Version: 0.3.0
+Version: 0.3.2
 Summary: SDK for the HUD platform.
-Project-URL: Homepage, https://github.com/hud-evals/hud-sdk
-Project-URL: Bug Tracker, https://github.com/hud-evals/hud-sdk/issues
+Project-URL: Homepage, https://github.com/hud-evals/hud-python
+Project-URL: Bug Tracker, https://github.com/hud-evals/hud-python/issues
 Project-URL: Documentation, https://docs.hud.so
 Author-email: HUD SDK <founders@hud.so>
 License: MIT License
@@ -35,28 +35,22 @@ Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Requires-Python: <3.14,>=3.11
-Requires-Dist: aiodocker>=0.24.0
 Requires-Dist: anthropic
+Requires-Dist: datasets>=4.0.0
 Requires-Dist: dotenv>=0.9.9
 Requires-Dist: httpx<1,>=0.23.0
-Requires-Dist: inspect-ai>=0.3.80
-Requires-Dist: ipykernel
 Requires-Dist: langchain
 Requires-Dist: langchain-anthropic
 Requires-Dist: langchain-openai
 Requires-Dist: mcp-use>=1.3.7
 Requires-Dist: mcp==1.12.2
-Requires-Dist: numpy
 Requires-Dist: openai
 Requires-Dist: pathspec>=0.12.1
-Requires-Dist: pillow>=11.1.0
-Requires-Dist: pyautogui>=0.9.54
 Requires-Dist: pydantic-settings<3,>=2
 Requires-Dist: pydantic<3,>=2
-Requires-Dist: textdistance<5,>=4.5.0
-Requires-Dist: toml>=0.10.2
 Requires-Dist: wrapt>=1.14.0
 Provides-Extra: dev
+Requires-Dist: aiodocker>=0.24.0; extra == 'dev'
 Requires-Dist: anthropic; extra == 'dev'
 Requires-Dist: dotenv; extra == 'dev'
 Requires-Dist: ipykernel; extra == 'dev'
@@ -64,17 +58,29 @@ Requires-Dist: ipython<9; extra == 'dev'
 Requires-Dist: jupyter-client; extra == 'dev'
 Requires-Dist: jupyter-core; extra == 'dev'
 Requires-Dist: openai; extra == 'dev'
+Requires-Dist: pillow>=11.1.0; extra == 'dev'
 Requires-Dist: playwright; extra == 'dev'
+Requires-Dist: pyautogui>=0.9.54; extra == 'dev'
 Requires-Dist: pyright==1.1.401; extra == 'dev'
 Requires-Dist: pytest-asyncio; extra == 'dev'
 Requires-Dist: pytest-cov; extra == 'dev'
 Requires-Dist: pytest-mock; extra == 'dev'
 Requires-Dist: pytest<9,>=8.1.1; extra == 'dev'
 Requires-Dist: ruff==0.11.8; extra == 'dev'
+Requires-Dist: toml>=0.10.2; extra == 'dev'
+Provides-Extra: v2
+Requires-Dist: aiodocker>=0.24.0; extra == 'v2'
+Requires-Dist: inspect-ai>=0.3.80; extra == 'v2'
+Requires-Dist: ipykernel; extra == 'v2'
+Requires-Dist: numpy; extra == 'v2'
+Requires-Dist: pillow>=11.1.0; extra == 'v2'
+Requires-Dist: pyautogui>=0.9.54; extra == 'v2'
+Requires-Dist: textdistance<5,>=4.5.0; extra == 'v2'
+Requires-Dist: toml>=0.10.2; extra == 'v2'
 Description-Content-Type: text/markdown
 
 <div align="left">
-  <img src="https://raw.githubusercontent.com/hud-evals/hud-sdk/main/docs/logo/hud_logo.svg" alt="HUD" width="150" style="margin-bottom: 20px;"/>
+  <img src="https://raw.githubusercontent.com/hud-evals/hud-python/main/docs/logo/hud_logo.svg" alt="HUD" width="150" style="margin-bottom: 20px;"/>
 </div>
 
 <h3>
@@ -88,7 +94,7 @@ Evaluate your Computer Use AI agents across web browsers, desktop environments,
 We're here to help with eval strategies, custom environments, or improving your agent architecture!
 
 
-> **Early Release Notice**: We'd love to hear your feedback in [Issues](https://github.com/hud-evals/hud-sdk/issues), as the SDK is still evolving!
+> **Early Release Notice**: We'd love to hear your feedback in [Issues](https://github.com/hud-evals/hud-python/issues), as the SDK is still evolving!
 
 [![PyPI version](https://img.shields.io/pypi/v/hud-python)](https://pypi.org/project/hud-python/)
 
@@ -272,7 +278,7 @@ If you use this SDK in your research, please cite it as follows:
   author = {HUD and Jay Ram and Lorenss Martinsons and Parth Patel and Oskars Putans and Govind Pimpale and Mayank Singamreddy and Nguyen Nhat Minh},
   title = {{HUD: An Evaluation Platform for Agents}},
   date = {2025-04},
-  url = {https://github.com/hud-evals/hud-sdk},
+  url = {https://github.com/hud-evals/hud-python},
   langid = {en}
 }
 ```

{hud_python-0.3.0.dist-info → hud_python-0.3.2.dist-info}/RECORD

@@ -1,23 +1,24 @@
-hud/__init__.py,sha256=yg4CC0iQWE67NGb6tUTmlO1kV-tM3njbigTuFYyzgAI,1477
+hud/__init__.py,sha256=j5Zzth7_M-5DU_KJT2ZV9OfikD2aE6lzyiZA4OrLzi8,1578
+hud/datasets.py,sha256=UZCzzXREbPhlw2ZdUFZ8EDz0lErWEeBPOPQxH71p6EA,6196
 hud/exceptions.py,sha256=Xna_pdEK_ESwkcffsRmT5GXq4xSHLV5cu7Qu3MjstSE,5516
-hud/gym.py,sha256=JNWlO2GXev0xIjoTI9HMEbcQgGpzc6fku7-RYoYAxHI,4996
+hud/gym.py,sha256=-hp5HdPBWf6-j0CgSoX_f2CTLssf1Wo5UhfyrnPbvkc,4774
 hud/job.py,sha256=0vWbr3E5bYstVRzXS_6l-57JGUFcrZpmFrNkOSQ8Aa0,26969
-hud/settings.py,sha256=rZFd_fzPUZKOklhMpTTSIJrMD1-eH9h6WrbD27SSXZ8,2014
-hud/task.py,sha256=WxftOrmaHqNvEsie1ZVIXJELYpfC9ejJL7b9TPQXEDg,8913
-hud/taskset.py,sha256=9IRwHeAdsk_IEibayM-hElE3gTp0mgmi-huN67h9-tc,7019
-hud/trajectory.py,sha256=ctAwrGIkdULr4xI6G-1Dp2fhDol4o_PmnPcqTzAEIUc,3797
+hud/settings.py,sha256=KPzeF9OUecApYH8YYMW-8vIRhFP_6htzzZvC4RCUARc,2183
+hud/task.py,sha256=l2mQM5Yc45kWjMXJkg1hVJfG0DLzTHAIXEvl4WLG-ho,5451
+hud/taskset.py,sha256=QjHbcxSy7h7fmtzRHW1ewxtOIydtH7ZotttDoiABTEY,6573
+hud/trajectory.py,sha256=LBVkFz6U_rmyooCZHN81tdOx0Z7DuAgzf0KQLejc4Fo,3937
 hud/types.py,sha256=h7fUowbdyGF4Fg8TUnvCFoa2fflRRPi6xx7YgpBwFis,3109
-hud/version.py,sha256=RPgsYJZ_E7U1ryhfdldTxD4xSiU4nQTEdzeq7iLynA4,104
+hud/version.py,sha256=xXGUzDnO0wgnaDf7cvjChhGymzp3vrfpGjE5wBibi8E,104
 hud/adapters/__init__.py,sha256=zz24KdC_e9TJPgWo6y57_8SzevEE5ak4Cm6tXzMxwRk,266
 hud/adapters/claude/__init__.py,sha256=i7QEF-29FLb9qxp1eYtXs-adIk_tG54tL-9g6d3xodk,100
 hud/adapters/claude/adapter.py,sha256=vCpotJ5gzQs4PP2iCXVavIcyG8c_4m1P6fuXStwUxSo,6675
 hud/adapters/claude/tests/__init__.py,sha256=9GZj0rz4tTkiPnLfxTmyBPr-s8UZc3gph6WH8fs8T34,39
 hud/adapters/claude/tests/test_adapter.py,sha256=cAdHEoqLngLiV7QwlWJ0KuNgb1vNv9WZTPQMnxhMDKI,18319
 hud/adapters/common/__init__.py,sha256=BjdZWJVs_AKtpFrt-tNsdQRjnz7D97DFEQirJ-r0mp8,118
-hud/adapters/common/adapter.py,sha256=GETzlsEl-uYkL-U4cQHBnfLAvm1dbXec4fKC2ypR1L0,5821
+hud/adapters/common/adapter.py,sha256=fTpw7wA501nxM3ufl6WMWq4Nc3vXlUeBGS7WgvZVFjU,6180
 hud/adapters/common/types.py,sha256=6frue7_gZlSYtOHhF2tFHqzjltzzHsTVs6-H-jQwZ4Y,9955
 hud/adapters/common/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-hud/adapters/common/tests/test_adapter.py,sha256=rTD36LjvytHqMIyOLDyrn0RLIkd20s6f6dwoBEarJaw,8744
+hud/adapters/common/tests/test_adapter.py,sha256=7QRpQPGM1PlMi8RcqJAT4ruGvLT9TgGmc9R5tzncN1M,8965
 hud/adapters/operator/__init__.py,sha256=31vTRs268_TOLd-TeQRKau5bDYy78wxCNpJFhD5_l8U,104
 hud/adapters/operator/adapter.py,sha256=Uz4Sr73T57B7v4RRP0uaibHI17N2hBx6Z9YYjgJCUXA,3732
 hud/adapters/operator/tests/__init__.py,sha256=yTsDVusVXZBQL6DnXpLgKQCBRuOYUAVQ8Blk_k5GETk,41
@@ -34,61 +35,55 @@ hud/agent/tests/__init__.py,sha256=HbAW7FvSvzzKPU5LpveZceU8XTcDkRe1Bmte3OGi2f0,2
 hud/agent/tests/test_base.py,sha256=MAHx4QWsX4y4jXDoA1sxWw8uFvL7lIzGlXrnHfOTmkw,8511
 hud/env/__init__.py,sha256=wVEesXMXM5hcNXQHt0-PN4-9RnE69DEnQENS7uJSv_Y,266
 hud/env/client.py,sha256=brhfLkWGSuvxl3vqGMCQT-vXfj8rUbJMhE3zJg9WMDA,869
-hud/env/docker_client.py,sha256=_EfxCbld2lk5BCBegBMMGXrYxOtxoa8N468T1wFbGrE,11234
+hud/env/docker_client.py,sha256=55PTFansUDzsRMT_43eSTVO9rb_wzl_s4aBpBqmMeXk,11749
 hud/env/environment.py,sha256=wjMBwGs5qkkXsVlXR_Z2QPZi4cwXE82ckdzRgHiXPjw,17019
-hud/env/local_docker_client.py,sha256=EbULGazP9KlD1tQrFKSghC0MO2-G60iNVLinEPtQ33M,11573
+hud/env/local_docker_client.py,sha256=IIuPSV_KJsfCONJAIVkgq_2zgUJl-FE4e5tDkkbRp0Y,12442
 hud/env/remote_client.py,sha256=tP5Gn1YtYgsjdXA4vM4FibAAHnR-9OOH4GrTog97cf8,6670
 hud/env/remote_docker_client.py,sha256=sBoOz3cq9HMgVvX8qCYEhRLvdswMZLG9G4Ybc60RzDo,9574
-hud/evaluators/__init__.py,sha256=V5nktEAw3EDn2Y537pjia5Y1IjdLBIPrDjTs6YTCdX4,153
-hud/evaluators/base.py,sha256=ALO9Rj-R_9HtHIHYp84bsQQD12De0XnCTwad78_T5-k,771
-hud/evaluators/inspect.py,sha256=ZvrTXLpgibyvQ5aNXAMP4quyXISrRQHg9besDcuCx7U,692
-hud/evaluators/judge.py,sha256=N3gEQGwVin9Ir80wWw6VtaL0xrlzitbmItaLm0he5gY,5962
-hud/evaluators/match.py,sha256=8YVQD942myX72Jkme2JFIVlmKhFXEa3CgGTjLC8O5n4,4701
-hud/evaluators/remote.py,sha256=kmD_XIU20KvX0NKgaEEKTTKHp0KVRa_3jUEgONh2nkY,2054
-hud/evaluators/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-hud/evaluators/tests/test_inspect.py,sha256=8dMjgQfXOJGcS8gP6TzoBbQiG_NYuRL6IobMG7euJdU,376
-hud/evaluators/tests/test_judge.py,sha256=c1GaAeq_WpBVgBlx-gQncHrOPokzKNxlbgiC8W8hxYI,7829
-hud/evaluators/tests/test_match.py,sha256=C04GoluyT9i41YZ65xEjN7tKHQbENbrpNhNtUd4ivmA,3919
-hud/evaluators/tests/test_remote.py,sha256=YdJpyyuRLkYP0e3jTUkD3zobS2WHQPePn8yBZtYOIN4,3243
-hud/mcp_agent/__init__.py,sha256=0R8SGgg2XU25y7B4lnBRv1n33d9TV6vaPXLafoiya2Y,324
-hud/mcp_agent/base.py,sha256=P92Bcj3VH8veWgG6Yrq6cnE2gOnRaVG0NhEXdI-C8CA,29142
-hud/mcp_agent/claude.py,sha256=5ORCs8PecqkRy2h5pVadxCIzJkjXZPPgkfOsGwJcJR4,11691
-hud/mcp_agent/langchain.py,sha256=JOD10jeFuW4ekgEu7fzKWuveBTTOV0CTIld98fNMbz0,8136
-hud/mcp_agent/openai.py,sha256=7SvbuKraLzlN4aGRsSkFtAVr1YldQmZ_9R8pRTWdQU0,12579
-hud/mcp_agent/tests/__init__.py,sha256=W-O-_4i34d9TTyEHV-O_q1Ai1gLhzwDaaPo02_TWQIY,34
-hud/mcp_agent/tests/test_base.py,sha256=7j_Id__Fd-d0VDRmfqyYM_p8JtF35mTPR90I8LeUXrI,16109
+hud/mcp/__init__.py,sha256=VBAZfpD8Ww59CkWb4CB0rGhNGqJYtc5y1gWZWHDaViQ,371
+hud/mcp/base.py,sha256=H4CRVGG4aEXAk_qRk3iOi-KLf8AVuffmoXPTaSXD4_0,24376
+hud/mcp/claude.py,sha256=XxXHjNnBvrS2Y98m0xTfFjZYgACCoFVTiNd01neffbM,12034
+hud/mcp/client.py,sha256=qrmpk2syjJ56y-09Dg44RVjUCFfmf5bPXaQSY-2ih-k,11494
+hud/mcp/langchain.py,sha256=hbKSCSQBf4W_pPpGEdy_KNoPA-T7Bsn_BLIDxaLzvVU,9251
+hud/mcp/openai.py,sha256=tpYK4ixLWqxAUXatXhoIZUXMlK1oP8TUZjnkSxBQVMc,13244
+hud/mcp/tests/__init__.py,sha256=W-O-_4i34d9TTyEHV-O_q1Ai1gLhzwDaaPo02_TWQIY,34
+hud/mcp/tests/test_base.py,sha256=lrRZoyDN7T67kOfv1A5WESaSHsYCaodD2CJnFli-4A4,19125
+hud/mcp/tests/test_claude.py,sha256=kGDThen8ij9QWx_YH3P9UvLlra1ueEMgA_clQ1q60II,11312
+hud/mcp/tests/test_client.py,sha256=ffxKzLmY75v-9l3aceUkn7aTdoO3j6deA4KBE3l9gaQ,11975
+hud/mcp/tests/test_openai.py,sha256=AhnBT_y-zMykQyJARDwKWiQWJsBGwNIlH6fGAzhJh88,9091
 hud/server/__init__.py,sha256=IPxPCqtPLguryN-nBq78Sakypw2bRiE2iHv3SXG8YRk,139
 hud/server/requests.py,sha256=AnFW4ELojjvfF6xjS2no6_fg4Rph2aR2hjPzYTede0Q,8841
 hud/server/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hud/server/tests/test_requests.py,sha256=63YCbykcib5MxKxm-OgHJPLX3QC7hmgIwnWaYukVM6s,9077
-hud/telemetry/__init__.py,sha256=tNRbewoND4DvZUNw2aRYpR4kjRuh4o1XoNUuis62Zes,558
-hud/telemetry/_trace.py,sha256=zggA2kTqLvM59F-DyRSRGKX-in27jcdrnr_xiTRWJvA,5375
-hud/telemetry/context.py,sha256=LyibWGCq4FhGzauxIws36Hzzd0S0ryMYCPqYwoot4DQ,4638
-hud/telemetry/exporter.py,sha256=LaD-U8TldKVn-rQkO88l8ZBlFxhVRbOuQmVKOGMMWnU,17875
+hud/telemetry/__init__.py,sha256=qSQhbXYy7c_sG7KhVr-5eiCmeREj6GQ2cijhbIR2-Z4,717
+hud/telemetry/_trace.py,sha256=Di2zKByHaljL6H4VkA-Gh_085jRJQw2VTiMOHX_FKp0,11433
+hud/telemetry/context.py,sha256=qwCdUQ3UX_Y_zfIHSAQ1cdJNv-VLh5y8ovXfLpjHKVY,7492
+hud/telemetry/exporter.py,sha256=10NwliO35J0fStvspgzb93N5MTko3pYNJe0fuTs-gPQ,23225
+hud/telemetry/job.py,sha256=eyjr7Ha2ijM0MIF5f0d1xFOScFUdFIqlmO8GzQZoAJc,4905
 hud/telemetry/mcp_models.py,sha256=0FQZoXtKOKeUsc2L61UbANpUDC7VNL842R2YFR61UBQ,8980
 hud/telemetry/instrumentation/__init__.py,sha256=vHmSqaJMMehgRNn6EN2SMoYDD12rSHkLeVmj7Uy1my0,88
 hud/telemetry/instrumentation/mcp.py,sha256=RbEaqmp8QHj1XqpIzwDSE8gH2cN5UjaBTouRxiPWxmc,9339
 hud/telemetry/instrumentation/registry.py,sha256=UVaSsEA693lvKYd5R3n3ve6GcAB1fwqubRwIVeZiNmo,1821
 hud/telemetry/tests/__init__.py,sha256=QMN8OzfrBUDbQESwrwHCqXLdDwCjYWX8BJcpeLUJfqA,33
 hud/telemetry/tests/test_context.py,sha256=RdtjYHsyvlkKoTQxk0VezaAISEoVQReYqQiqK3jgFLQ,6746
-hud/telemetry/tests/test_trace.py,sha256=fZt8WXflZivhBgWHhePWmmNbTYg0qF3oF3bTKE0KXiM,12016
-hud/tools/__init__.py,sha256=vDxoIGHaj7FOnqki2Q92gGmEZx3f6Vx9RGkQo-X3cJ4,577
+hud/telemetry/tests/test_trace.py,sha256=mCm5AH-NpuDVvRG-CZhMMqHiJ4dahvcy9KHmWmo6o3A,12494
+hud/tools/__init__.py,sha256=T4PnE5nuBCXsTKXUYBHmaF1Ojc6D5vAa6wA2cFWJfTc,986
 hud/tools/base.py,sha256=lmd7N7IccIWrPpA0NZundIglFTTiLFW9VP_PJI2EXug,2069
 hud/tools/bash.py,sha256=o841_HF1NJFfUWLOVUw9s0iB4BoIxhA-8vMasJOhZ70,4319
 hud/tools/edit.py,sha256=9vJ2XSnWOPViujQbZZuDjLahvzxoPHyAeXxgKfpUDHo,11796
-hud/tools/playwright_tool.py,sha256=tq1La66esH2CwGYBkpvBDNPuswsspHtSE1cSqGVJEjU,13295
+hud/tools/playwright_tool.py,sha256=IQT1hk5U4H8BI988iZq0B2oS_fbgkaX01Z-ZXL4r71o,13724
 hud/tools/utils.py,sha256=bfVyYMcBOJvr1QdptCjVb6jaHVGIL5WUxmY59kzMekQ,1447
 hud/tools/computer/__init__.py,sha256=ehKY7u0_4cZ9h7YQlOQjbKPWfd5LhQq8ZQn2w2-l2mY,302
 hud/tools/computer/anthropic.py,sha256=M-djQmd0vPZm95FDszaMh4wSaLFPhlcCUb-JkSuflnU,16104
-hud/tools/computer/hud.py,sha256=xyFYLqVoLsps0Dbs9kAfg941kXLnMHx7SL8a2skhjHw,13351
+hud/tools/computer/hud.py,sha256=13_xjvf5-yO-7lYBoI44Br31CxL5EumSCQxq876h7rs,13840
 hud/tools/computer/openai.py,sha256=pcMGfoT6O8Rh9IrW_H1Mw2cIwk-FzCswrgjW19piRU8,10538
-hud/tools/executors/__init__.py,sha256=Ybc8mP48ps3Z2QHjYcc0Yrhmn2ZNqZF1WLvl-0lyQ_w,262
+hud/tools/executors/__init__.py,sha256=jHxfus9SLhkL6YGtebR5RyKYyVAix3yu5EkUp2Q27Kg,732
 hud/tools/executors/base.py,sha256=4h04Byt4ktaNk_aLOOI798pkMCLiqA7pE2PoaEn_hfg,11647
-hud/tools/executors/pyautogui.py,sha256=1MWWXhyaPLeFkWXIr7pR_pii_XjDJxhpXdCCHFgF-1A,20803
+hud/tools/executors/pyautogui.py,sha256=Kc2OcFw-sEuRBRFtO1ZrWeHs1p-p5FtEpESkzpRhOHk,22098
 hud/tools/executors/xdo.py,sha256=C6ecIVPUba7c6vKpgIcNxKcc698hwelQjj4YYUxT2_4,17751
 hud/tools/executors/tests/__init__.py,sha256=opFpGSH6cEqIZgt9izXd3Yt85pC7xkxiYmOZQTHf4AY,32
 hud/tools/executors/tests/test_base_executor.py,sha256=dvpKHCIjrBhT6E2U3hsjAwuivCAYXplvd08EHN6cxTI,12306
-hud/tools/executors/tests/test_pyautogui_executor.py,sha256=cBgTAieWVT6C9dRtdqBWtRFmayteVQEp6DNofdCYLqc,6521
+hud/tools/executors/tests/test_pyautogui_executor.py,sha256=br-wVvXnRx9G6X0yJ_xeKZf2xl8o4LCnYLeaIbkpuzY,6608
 hud/tools/helper/README.md,sha256=GDS-K-wMnDO3-gtWjisgk5153zBmU29XSrs2ZhlOWQY,1727
 hud/tools/helper/__init__.py,sha256=VqgQkY-y9h-WnGXZRK387fSr1BzrOQoAy3975WDAs4c,209
 hud/tools/helper/mcp_server.py,sha256=t8UaGq91hDKef6zO3ApnJydwcKEqgLF6RdDcJ1GmfEA,2248
@@ -98,10 +93,10 @@ hud/tools/tests/__init__.py,sha256=eEYYkxX5Hz9woXVOBJ2H2_CQoEih0vH6nRt3sH2Z8v8,4
 hud/tools/tests/test_bash.py,sha256=LV3LjijwkQqxuxIXFSepD2x3sYoY4uhdw8EBv4JOyLU,4847
 hud/tools/tests/test_computer.py,sha256=HxYHxKJ0eWyZzC3abzviFBU-auc8x6Sh2ciR_uVXMXw,1595
 hud/tools/tests/test_computer_actions.py,sha256=YtUNFL7anhpXrcvg8EoUY1CqIV-TAAyaNFLZO9CiJ40,1194
-hud/tools/tests/test_edit.py,sha256=UpUkn-fEXyFr9dKPT7pjmZ8ASUePkPnqwVMplUHowR4,8301
+hud/tools/tests/test_edit.py,sha256=_Bfh9Qc_zSYK5vS9kfhm5G9tkVvX1dsEIFqE3jkeSv0,8527
 hud/tools/tests/test_init.py,sha256=PD_SS6X6SPhEjStJqYxdJRtsa7RbL6cTokAGIn5bWhA,702
 hud/tools/tests/test_playwright_tool.py,sha256=1qED_NF2QXUZmBRbWSmcKImMLUQ3m5CbA_9tLUiaxTQ,6696
-hud/tools/tests/test_tools.py,sha256=fNZMv93VhkFiXRHKt5krXZAASU2IsiA1149erUMa2ek,4418
+hud/tools/tests/test_tools.py,sha256=KgSPgdqldpifbHeQHBFdYJVf3boWbvK6LRRRORPfTOg,4595
 hud/tools/tests/test_utils.py,sha256=oYxEnLpSA5sEeYFGUTj74QRNv0AHP3AjmYYHXgIW0BY,5496
 hud/utils/__init__.py,sha256=oSl_gGoS272X2VFnBYX8hLxcP2xgGoBYQXAuLhtQgw8,260
 hud/utils/agent.py,sha256=CpNgjKWMaNqo-EATH_vfJHIN53rEkZngm2LXfUFlldQ,1225
@@ -116,9 +111,9 @@ hud/utils/tests/test_config.py,sha256=dPlXYWuMrxX-NOYbf0vdJ27TJpfacKG8eiKOSGOcfD
 hud/utils/tests/test_init.py,sha256=UxlNTwjlSE2q3M0R86EmMYmmXmbRvzZaC-S2av26QXI,529
 hud/utils/tests/test_progress.py,sha256=QunwDgi_heQXhDgmC25zgjr-sFUu5FdJ_1aYigMKeIc,6351
 hud/utils/tests/test_telemetry.py,sha256=t0An1RTBaE0dZVEpF4uwuq5k1R-PXFR5k4u71h60tx8,1224
-hud/utils/tests/test_version.py,sha256=J-6wPtkqEIDGjTCmUIylXicun-t7JdEHaJ8gcuj2mlY,159
+hud/utils/tests/test_version.py,sha256=b0JOKjG9oz4PvBZxUS3hbQYWi8yTEz52VHqTlLXgSYM,159
 hud/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-hud_python-0.3.0.dist-info/METADATA,sha256=fgbwnVd2CqZ4TSYMjjj9ygmhG3Yu_UnahOlqPXMcRQ4,9876
-hud_python-0.3.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-hud_python-0.3.0.dist-info/licenses/LICENSE,sha256=yIzBheVUf86FC1bztAcr7RYWWNxyd3B-UJQ3uddg1HA,1078
-hud_python-0.3.0.dist-info/RECORD,,
+hud_python-0.3.2.dist-info/METADATA,sha256=p5VyDJbUHnHXC1cUh50DJnVF8j8NIFqjEi-bdd8vP8c,10249
+hud_python-0.3.2.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+hud_python-0.3.2.dist-info/licenses/LICENSE,sha256=yIzBheVUf86FC1bztAcr7RYWWNxyd3B-UJQ3uddg1HA,1078
+hud_python-0.3.2.dist-info/RECORD,,

hud/evaluators/__init__.py

@@ -1,9 +0,0 @@
-"""
-Evaluators for assessing task responses.
-"""
-
-from __future__ import annotations
-
-from hud.evaluators.base import Evaluator
-
-__all__ = ["Evaluator"]

hud/evaluators/base.py

@@ -1,32 +0,0 @@
-from __future__ import annotations
-
-from abc import ABC, abstractmethod
-from typing import TYPE_CHECKING
-
-from pydantic import BaseModel, Field
-
-if TYPE_CHECKING:
-    from hud.task import Task
-
-
-class EvaluationResult(BaseModel):
-    """Result of an evaluation.
-
-    Attributes:
-        score: Float score between 0 and 1
-        reason: Explanation of the evaluation
-        mode: Mode used for matching, if applicable
-    """
-
-    score: float
-    reason: str
-    mode: str | None = None
-    criteria_scores: dict[str, float] | None = Field(default_factory=dict)
-
-
-class Evaluator(ABC):
-    """Abstract base class for evaluators."""
-
-    @abstractmethod
-    def evaluate(self, task: Task, response: str) -> EvaluationResult:
-        """Evaluate a task and response."""

hud/evaluators/inspect.py

@@ -1,24 +0,0 @@
-from __future__ import annotations
-
-from typing import Any
-
-from hud.evaluators.base import EvaluationResult
-
-
-def inspect_evaluate(
-    response: Any,
-    answer: Any,
-) -> EvaluationResult:
-    """Evaluate using Inspect-ai's evaluation models.
-
-    Args:
-        response: The response to evaluate
-        answer: The reference answer to compare against
-        model_name: The Inspect model to use
-        prompt: Optional custom prompt for evaluation
-        metrics: Optional list of metrics to evaluate against
-
-    Returns:
-        EvaluationResult with the evaluation results
-    """
-    return EvaluationResult(score=0.0, reason="Inspect evaluation not implemented", mode="inspect")

hud/evaluators/judge.py

@@ -1,189 +0,0 @@
-from __future__ import annotations
-
-import asyncio
-import base64
-from typing import Any, Protocol, TypedDict
-
-from hud.evaluators.base import EvaluationResult
-from hud.server import make_request
-from hud.settings import settings
-
-
-class LLM(Protocol):
-    """Protocol for LLM interfaces that can be used for evaluation."""
-
-    async def ainvoke(self, prompt: str, /) -> str: ...
-
-
-class Criterion(TypedDict, total=False):
-    """Criterion for judge-based evaluation."""
-
-    description: str
-    weight: float
-
-
-async def _call_eval_endpoint(
-    response: Any, answer: Any, criteria: list[Any], mode: str
-) -> dict[str, Any]:
-    """Call the run_eval endpoint to evaluate the response."""
-    try:
-        result = await make_request(
-            method="POST",
-            url=f"{settings.base_url}/evaluations/run_eval",
-            json={"response": response, "answer": answer, "criteria": criteria, "mode": mode},
-            api_key=settings.api_key,
-        )
-        return result
-    except Exception as e:
-        # Fallback to local evaluation if remote call fails
-        return {
-            "score": -1.0,
-            "reason": f"Remote evaluation failed: {e!s}. Fallback to default score.",
-            "criteria_scores": {},
-        }
-
-
-def _process_input(data: Any) -> Any:
-    """Process input data, detecting and handling base64 images."""
-    if isinstance(data, bytes):
-        # Convert bytes to base64 string
-        return base64.b64encode(data).decode("utf-8")
-
-    if isinstance(data, str) and _is_base64_image(data):
-        # It's already a base64 string, just return it
-        return data
-
-    if isinstance(data, list) and all(isinstance(item, str) for item in data):
-        # Process list of strings
-        return data
-
-    # For other types, convert to string
-    return str(data) if not isinstance(data, str | dict) else data
-
-
-def _is_base64_image(data: Any) -> bool:
-    """Check if a string is a base64 encoded image."""
-    if not isinstance(data, str):
-        return False
-
-    # Check for common image data URI pattern
-    if data.startswith(("data:image/", "data:application/octet-stream")):
-        return True
-
-    # Check if it's a base64 encoded string with image header
-    try:
-        # First, validate it's base64 decodable
-        padding_needed = len(data) % 4
-        if padding_needed:
-            data += "=" * (4 - padding_needed)
-
-        # Try to decode the first few bytes to check for image signatures
-        sample = base64.b64decode(data[:30])
-
-        # Check for common image format signatures
-        return sample.startswith((b"\xff\xd8\xff", b"\x89PNG\r\n\x1a\n", b"GIF8", b"RIFF"))
-    except Exception:
-        return False
-
-
-def judge(
-    response: Any,
-    answer: Any,
-    llm: LLM | None = None,
-    criteria: list[str] | list[dict] | None = None,
-) -> EvaluationResult:
-    """Judge a response against an answer using an LLM.
-
-    Args:
-        response: The response to evaluate
-        answer: The reference answer to compare against
-        llm: Optional langchain LLM to use for evaluation
-        criteria: Evaluation criteria as strings or dictionaries
-
-    Returns:
-        EvaluationResult with evaluation results
-    """
-    # Process inputs
-    processed_response = _process_input(response)
-    processed_answer = _process_input(answer)
-
-    # If LLM is provided, use it for evaluation
-    if llm:
-        return _evaluate_with_llm(processed_response, processed_answer, llm, criteria)
-
-    # Otherwise, use the remote evaluation service
-    mode = "LLM"
-    if isinstance(answer, bytes) or _is_base64_image(answer):
-        mode = "VLM"
-
-    # Call the eval endpoint synchronously
-    result = asyncio.run(
-        _call_eval_endpoint(
-            response=processed_response, answer=processed_answer, criteria=criteria or [], mode=mode
-        )
-    )
-
-    return EvaluationResult(
-        score=result.get("score", -1.0),
-        reason=result.get("reason", "Response evaluated"),
-        mode=mode,
-        criteria_scores=result.get("criteria_scores", {}),
-    )
-
-
-def _evaluate_with_llm(
-    response: Any, answer: Any, llm: LLM, criteria: list[str] | list[dict] | None = None
-) -> EvaluationResult:
-    """Evaluate a response against an answer using a provided LLM."""
-    criteria_text = ""
-    if criteria:
-        criteria_text = "Use the following criteria:\n"
-        for c in criteria:
-            if isinstance(c, dict) and "description" in c:
-                criteria_text += f"- {c['description']}\n"
-            elif isinstance(c, str):
-                criteria_text += f"- {c}\n"
-
-    prompt = f"""Evaluate the quality of a response given a reference answer.
-
-REFERENCE ANSWER:
-{answer}
-
-RESPONSE TO EVALUATE:
-{response}
-
-{criteria_text}
-Rate the response on a scale from 0.0 to 1.0, where 1.0 is perfect.
-Provide a brief explanation for your rating.
-Format your answer as a JSON object with 'score' (float) and 'reason' (string) fields.
-"""
-
-    try:
-        # Run the evaluation asynchronously
-        result_text = asyncio.run(llm.ainvoke(prompt))
-
-        # Attempt to parse JSON response
-        import json
-        import re
-
-        # Try to extract JSON if wrapped in other text
-        json_match = re.search(r"\{.*?\}", result_text, re.DOTALL)
-        if json_match:
-            json_str = json_match.group(0)
-            result = json.loads(json_str)
-
-            return EvaluationResult(
-                score=float(result.get("score", 0.5)),
-                reason=result.get("reason", "Evaluated with custom LLM"),
-                mode="custom_llm",
-            )
-
-        # If can't parse as JSON, use default values
-        return EvaluationResult(
-            score=0.5,
-            reason=f"Unable to parse LLM response as JSON. Raw response: {result_text[:100]}...",
-            mode="custom_llm",
-        )
-
-    except Exception as e:
-        return EvaluationResult(score=0.0, reason=f"LLM evaluation error: {e!s}", mode="custom_llm")

hud/evaluators/match.py

@@ -1,156 +0,0 @@
-from __future__ import annotations
-
-import re
-from difflib import SequenceMatcher
-from typing import TYPE_CHECKING, Protocol
-
-from textdistance import levenshtein
-
-from hud.evaluators.base import EvaluationResult
-
-if TYPE_CHECKING:
-    from collections.abc import Sequence
-
-
-class _Stringable(Protocol):
-    def __str__(self) -> str: ...
-
-
-def match_single(response: _Stringable, answer: _Stringable) -> EvaluationResult:
-    """Check if the answer is present within the response.
-
-    Args:
-        response: The response to evaluate
-        answer: The expected answer
-
-    Returns:
-        EvaluationResult with score=1.0 if match, 0.0 otherwise
-    """
-    passed = str(answer).lower().strip() in str(response).lower().strip()
-    return EvaluationResult(
-        score=1.0 if passed else 0.0,
-        reason="Exact match" if passed else "No exact match found",
-        mode="single",
-    )
-
-
-def match_all(response: _Stringable, answers: Sequence[_Stringable]) -> EvaluationResult:
-    """Count how many expected answers are in the response.
-
-    Args:
-        response: The response to evaluate
-        answers: List of expected answers
-
-    Returns:
-        EvaluationResult with score=proportion of matches (0.0-1.0)
-    """
-    response_str = str(response).lower()
-    matches = 0
-
-    for answer in answers:
-        if str(answer).lower() in response_str:
-            matches += 1
-
-    score = matches / len(answers) if answers else 0.0
-
-    if matches == len(answers):
-        reason = f"All {matches} expected items found"
-    else:
-        reason = f"Only {matches} of {len(answers)} expected items found"
-
-    return EvaluationResult(score=score, reason=reason, mode="all")
-
-
-def match_fuzzy(response: _Stringable, answer: _Stringable) -> EvaluationResult:
-    """Calculate similarity using Levenshtein distance.
-
-    Args:
-        response: The response to evaluate
-        answer: The expected answer
-
-    Returns:
-        EvaluationResult with score=similarity (0.0-1.0)
-    """
-    s1 = str(response).lower()
-    s2 = str(answer).lower()
-
-    if s1 == s2:
-        score = 1.0
-    elif len(s1) == 0 or len(s2) == 0:
-        score = 0.0
-    else:
-        # Use Levenshtein distance
-        distance = levenshtein.distance(s1, s2)
-        max_len = max(len(s1), len(s2))
-        score = 1.0 - (distance / max_len)
-
-    return EvaluationResult(
-        score=score, reason=f"Fuzzy match with {score:.1%} similarity", mode="fuzz"
-    )
-
-
-def match_regex(response: _Stringable, pattern: str) -> EvaluationResult:
-    """Check if response matches regex pattern.
-
-    Args:
-        response: The response to evaluate
-        pattern: Regular expression pattern to match
-
-    Returns:
-        EvaluationResult with score=1.0 if match, 0.0 otherwise
-    """
-    try:
-        regex = re.compile(pattern, re.DOTALL)
-        passed = bool(regex.search(str(response)))
-        return EvaluationResult(
-            score=1.0 if passed else 0.0,
-            reason="Regex pattern matched" if passed else "Regex pattern did not match",
-            mode="regex",
-        )
-    except re.error:
-        return EvaluationResult(score=0.0, reason="Invalid regex pattern", mode="regex")
-
-
-def match_diff(response: _Stringable, answer: _Stringable) -> EvaluationResult:
-    """Compare difference between response and answer.
-
-    Args:
-        response: The response to evaluate
-        answer: The expected answer
-
-    Returns:
-        EvaluationResult with score=similarity (0.0-1.0)
-    """
-    if isinstance(response, int | float) and isinstance(answer, int | float):
-        score = _match_numeric_diff(response, answer)
-        reason = f"Numeric difference: {abs(response - answer)}"
-    else:
-        score = _match_string_diff(response, answer)
-        reason = f"String difference with {score:.1%} similarity"
-
-    return EvaluationResult(score=score, reason=reason, mode="diff")
-
-
-def _match_string_diff(response: _Stringable, answer: _Stringable) -> float:
-    """Compare difference between response and answer strings."""
-    matcher = SequenceMatcher(None, str(response), str(answer))
-    return matcher.ratio()
-
-
-def _match_numeric_diff(response: float, answer: float) -> float:
-    """Calculate normalized difference between numeric values.
-
-    Returns a value between 0 and 1, where 1 means identical and 0 means maximum difference.
-    """
-    if response == answer:
-        return 1.0
-
-    # Simple absolute difference normalized to a 0-1 scale
-    diff = abs(response - answer)
-    max_val = max(abs(response), abs(answer))
-
-    if max_val == 0:
-        return 1.0  # Both are zero
-
-    # Normalize and invert so 1.0 means identical
-    return max(0.0, 1.0 - min(1.0, diff / max_val))