ursa-ai 0.2.8__py3-none-any.whl → 0.2.10__py3-none-any.whl

ursa/agents/__init__.py

@@ -1,9 +1,17 @@
-from .planning_agent import PlanningAgent, PlanningState
-from .websearch_agent import WebSearchAgent, WebSearchState
-from .execution_agent import ExecutionAgent, ExecutionState
-from .code_review_agent import CodeReviewAgent, CodeReviewState
-from .hypothesizer_agent import HypothesizerAgent, HypothesizerState
-from .arxiv_agent import ArxivAgent, PaperState, PaperMetadata
-from .recall_agent import RecallAgent
-from .base import BaseAgent, BaseChatModel
-from .mp_agent import MaterialsProjectAgent
+from .arxiv_agent import ArxivAgent as ArxivAgent
+from .arxiv_agent import PaperMetadata as PaperMetadata
+from .arxiv_agent import PaperState as PaperState
+from .base import BaseAgent as BaseAgent
+from .base import BaseChatModel as BaseChatModel
+from .code_review_agent import CodeReviewAgent as CodeReviewAgent
+from .code_review_agent import CodeReviewState as CodeReviewState
+from .execution_agent import ExecutionAgent as ExecutionAgent
+from .execution_agent import ExecutionState as ExecutionState
+from .hypothesizer_agent import HypothesizerAgent as HypothesizerAgent
+from .hypothesizer_agent import HypothesizerState as HypothesizerState
+from .mp_agent import MaterialsProjectAgent as MaterialsProjectAgent
+from .planning_agent import PlanningAgent as PlanningAgent
+from .planning_agent import PlanningState as PlanningState
+from .recall_agent import RecallAgent as RecallAgent
+from .websearch_agent import WebSearchAgent as WebSearchAgent
+from .websearch_agent import WebSearchState as WebSearchState

ursa/agents/arxiv_agent.py

@@ -1,29 +1,29 @@
+import base64
 import os
-import pymupdf
-import requests
-import feedparser
-from PIL import Image
+import re
+import statistics
+from concurrent.futures import ThreadPoolExecutor, as_completed
 from io import BytesIO
-import base64
 from urllib.parse import quote
-from typing_extensions import TypedDict, List
-from concurrent.futures import ThreadPoolExecutor, as_completed
-from tqdm import tqdm
-import statistics
-import re
 
+import feedparser
+import pymupdf
+import requests
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_chroma import Chroma
 from langchain_community.document_loaders import PyPDFLoader
 from langchain_core.output_parsers import StrOutputParser
 from langchain_core.prompts import ChatPromptTemplate
-from langgraph.graph import StateGraph, END, START
-from langchain.text_splitter import RecursiveCharacterTextSplitter
-from langchain_chroma import Chroma
+from langgraph.graph import StateGraph
+from PIL import Image
+from tqdm import tqdm
+from typing_extensions import List, TypedDict
 
 from .base import BaseAgent
 
 try:
     from openai import OpenAI
-except:
+except Exception:
     pass
 
 # embeddings = GoogleGenerativeAIEmbeddings(model="models/embedding-001")
@@ -162,7 +162,7 @@ class ArxivAgent(BaseAgent):
                 full_id = entry.id.split("/abs/")[-1]
                 arxiv_id = full_id.split("/")[-1]
                 title = entry.title.strip()
-                authors = ", ".join(author.name for author in entry.authors)
+                # authors = ", ".join(author.name for author in entry.authors)
                 pdf_url = f"https://arxiv.org/pdf/{full_id}.pdf"
                 pdf_filename = os.path.join(
                     self.database_path, f"{arxiv_id}.pdf"
@@ -211,12 +211,10 @@ class ArxivAgent(BaseAgent):
                 except Exception as e:
                     full_text = f"Error loading paper: {e}"
 
-            papers.append(
-                {
-                    "arxiv_id": arxiv_id,
-                    "full_text": full_text,
-                }
-            )
+            papers.append({
+                "arxiv_id": arxiv_id,
+                "full_text": full_text,
+            })
 
         return papers
 
@@ -279,28 +277,23 @@ class ArxivAgent(BaseAgent):
                     )
 
                     if relevant_docs_with_scores:
-                        score = sum(
-                            [s for _, s in relevant_docs_with_scores]
-                        ) / len(relevant_docs_with_scores)
+                        score = sum([
+                            s for _, s in relevant_docs_with_scores
+                        ]) / len(relevant_docs_with_scores)
                         relevancy_scores[i] = abs(1.0 - score)
                     else:
                         relevancy_scores[i] = 0.0
 
-                    retrieved_content = "\n\n".join(
-                        [
-                            doc.page_content
-                            for doc, _ in relevant_docs_with_scores
-                        ]
-                    )
+                    retrieved_content = "\n\n".join([
+                        doc.page_content for doc, _ in relevant_docs_with_scores
+                    ])
                 else:
                     retrieved_content = cleaned_text
 
-                summary = chain.invoke(
-                    {
-                        "retrieved_content": retrieved_content,
-                        "context": state["context"],
-                    }
-                )
+                summary = chain.invoke({
+                    "retrieved_content": retrieved_content,
+                    "context": state["context"],
+                })
 
             except Exception as e:
                 summary = f"Error summarizing paper: {e}"
@@ -313,7 +306,7 @@ class ArxivAgent(BaseAgent):
 
         if "papers" not in state or len(state["papers"]) == 0:
             print(
-                f"No papers retrieved - bad query or network connection to ArXiv?"
+                "No papers retrieved - bad query or network connection to ArXiv?"
             )
             return {**state, "summaries": None}
 
@@ -376,9 +369,10 @@ class ArxivAgent(BaseAgent):
 
         chain = prompt | self.llm | StrOutputParser()
 
-        final_summary = chain.invoke(
-            {"Summaries": combined, "context": state["context"]}
-        )
+        final_summary = chain.invoke({
+            "Summaries": combined,
+            "context": state["context"],
+        })
 
         with open(self.summaries_path + "/final_summary.txt", "w") as f:
             f.write(final_summary)
@@ -406,9 +400,10 @@ class ArxivAgent(BaseAgent):
         return graph
 
     def run(self, arxiv_search_query: str, context: str) -> str:
-        result = self.graph.invoke(
-            {"query": arxiv_search_query, "context": context}
-        )
+        result = self.graph.invoke({
+            "query": arxiv_search_query,
+            "context": context,
+        })
 
         if self.summarize:
             return result.get("final_summary", "No summary generated.")

ursa/agents/base.py

@@ -1,9 +1,7 @@
 from langchain_core.language_models.chat_models import BaseChatModel
+from langchain_core.load import dumps
 from langchain_litellm import ChatLiteLLM
 from langgraph.checkpoint.base import BaseCheckpointSaver
-from langchain_core.load import dumps
-
-import json
 
 
 class BaseAgent:

ursa/agents/code_review_agent.py

@@ -7,7 +7,7 @@ from langchain_core.messages import HumanMessage, SystemMessage
 from langchain_core.tools import tool
 from langgraph.graph import END, START, StateGraph
 from langgraph.graph.message import add_messages
-from langgraph.prebuilt import ToolNode, InjectedState
+from langgraph.prebuilt import InjectedState, ToolNode
 from typing_extensions import TypedDict
 
 from ..prompt_library.code_review_prompts import (
@@ -349,7 +349,7 @@ def main():
     }
     result = (
         code_review_agent.action.invoke(initial_state),
-        {"configurable": {"thread_id": self.thread_id}},
+        {"configurable": {"thread_id": 42}},
     )
     for x in result["messages"]:
         print(x.content)

ursa/agents/execution_agent.py

@@ -107,7 +107,7 @@ class ExecutionAgent(BaseAgent):
             # note that we've done the symlink now, so don't need to do it later
             new_state["symlinkdir"]["is_linked"] = True
 
-        if type(new_state["messages"][0]) == SystemMessage:
+        if isinstance(new_state["messages"][0], SystemMessage):
             new_state["messages"][0] = SystemMessage(
                 content=self.executor_prompt
             )
@@ -139,7 +139,7 @@ class ExecutionAgent(BaseAgent):
             memories = []
             # Handle looping through the messages
             for x in state["messages"]:
-                if not type(x) == AIMessage:
+                if not isinstance(x, AIMessage):
                     memories.append(x.content)
                 elif not x.tool_calls:
                     memories.append(x.content)
@@ -426,7 +426,7 @@ def edit_code(
 
     if old_code_clean not in content:
         console.print(
-            f"[yellow] ⚠️ 'old_code' not found in file'; no changes made.[/]"
+            "[yellow] ⚠️ 'old_code' not found in file'; no changes made.[/]"
         )
         return f"No changes made to {filename}: 'old_code' not found in file."
 
@@ -483,7 +483,7 @@ def command_safe(state: ExecutionState) -> Literal["safe", "unsafe"]:
     index = -1
     message = state["messages"][index]
     # Loop through all the consecutive tool messages in reverse order
-    while type(message) == ToolMessage:
+    while isinstance(message, ToolMessage):
         if "[UNSAFE]" in message.content:
             return "unsafe"
 

ursa/agents/hypothesizer_agent.py

@@ -96,7 +96,7 @@ class HypothesizerAgent(BaseAgent):
             new_state["visited_sites"] = []
 
         try:
-            if type(raw_search_results) == str:
+            if isinstance(raw_search_results, str):
                 results_list = ast.literal_eval(raw_search_results)
             else:
                 results_list = raw_search_results
@@ -154,7 +154,7 @@ class HypothesizerAgent(BaseAgent):
             new_state["visited_sites"] = []
 
         try:
-            if type(raw_search_results) == str:
+            if isinstance(raw_search_results, str):
                 results_list = ast.literal_eval(raw_search_results)
             else:
                 results_list = raw_search_results
@@ -217,7 +217,7 @@ class HypothesizerAgent(BaseAgent):
             new_state["visited_sites"] = []
 
         try:
-            if type(raw_search_results) == str:
+            if isinstance(raw_search_results, str):
                 results_list = ast.literal_eval(raw_search_results)
             else:
                 results_list = raw_search_results
@@ -589,7 +589,7 @@ if __name__ == "__main__":
         initial_state,
         {
             "recursion_limit": 999999,
-            "configurable": {"thread_id": self.thread_id},
+            "configurable": {"thread_id": 42},
         },
     )
     summary_text = result["summary_report"]

ursa/agents/mp_agent.py

@@ -1,20 +1,15 @@
-import os
 import json
-from typing import List, Dict
-from concurrent.futures import ThreadPoolExecutor
-from tqdm import tqdm
-
-from mp_api.client import MPRester
-
 import os
-from typing_extensions import TypedDict, List
-from concurrent.futures import ThreadPoolExecutor
-from tqdm import tqdm
 import re
+from concurrent.futures import ThreadPoolExecutor
+from typing import Dict
 
 from langchain_core.output_parsers import StrOutputParser
 from langchain_core.prompts import ChatPromptTemplate
-from langgraph.graph import StateGraph, END, START
+from langgraph.graph import StateGraph
+from mp_api.client import MPRester
+from tqdm import tqdm
+from typing_extensions import List, TypedDict
 
 from .base import BaseAgent
 
@@ -146,9 +141,10 @@ You are a materials-science assistant. Given the following metadata about a mate
         {context}
                 """)
         chain = prompt | self.llm | StrOutputParser()
-        final = chain.invoke(
-            {"summaries": combined, "context": state["context"]}
-        )
+        final = chain.invoke({
+            "summaries": combined,
+            "context": state["context"],
+        })
         return {**state, "final_summary": final}
 
     def _build_graph(self):

ursa/agents/planning_agent.py

@@ -40,7 +40,7 @@ class PlanningAgent(BaseAgent):
 
     def generation_node(self, state: PlanningState) -> PlanningState:
         messages = state["messages"]
-        if type(messages[0]) == SystemMessage:
+        if isinstance(messages[0], SystemMessage):
             messages[0] = SystemMessage(content=self.planner_prompt)
         else:
             messages = [SystemMessage(content=self.planner_prompt)] + messages

ursa/agents/recall_agent.py

@@ -1,4 +1,3 @@
-import os
 from .base import BaseAgent
 
 

ursa/agents/websearch_agent.py

@@ -1,5 +1,3 @@
-import inspect
-
 # from langchain_community.tools    import TavilySearchResults
 # from langchain_core.runnables.graph import MermaidDrawMethod
 from typing import Annotated, Any, List, Optional
@@ -9,7 +7,6 @@ from bs4 import BeautifulSoup
 from langchain_community.tools import DuckDuckGoSearchResults
 from langchain_core.language_models import BaseChatModel
 from langchain_core.messages import HumanMessage, SystemMessage, ToolMessage
-from langchain_core.tools import tool
 from langchain_openai import ChatOpenAI
 from langgraph.graph import END, START, StateGraph
 from langgraph.graph.message import add_messages
@@ -19,8 +16,8 @@ from typing_extensions import TypedDict
 
 from ..prompt_library.websearch_prompts import (
     reflection_prompt,
-    websearch_prompt,
     summarize_prompt,
+    websearch_prompt,
 )
 from .base import BaseAgent
 
@@ -41,6 +38,7 @@ class WebSearchState(TypedDict):
     remaining_steps: int
     is_last_step: bool
     model: Any
+    thread_id: Any
 
 
 # Adding the model to the state clumsily so that all "read" sources arent in the
@@ -113,8 +111,14 @@ class WebSearchAgent(BaseAgent):
         except (requests.ConnectionError, requests.Timeout):
             return False
 
+    def state_store_node(self, state: WebSearchState) -> WebSearchState:
+        state["thread_id"] = self.thread_id
+        return state
+        # return dict(**state, thread_id=self.thread_id)
+
     def _initialize_agent(self):
         self.graph = StateGraph(WebSearchState)
+        self.graph.add_node("state_store", self.state_store_node)
         self.graph.add_node(
             "websearch",
             create_react_agent(
@@ -128,7 +132,8 @@ class WebSearchAgent(BaseAgent):
         self.graph.add_node("review", self.review_node)
         self.graph.add_node("response", self.response_node)
 
-        self.graph.add_edge(START, "websearch")
+        self.graph.add_edge(START, "state_store")
+        self.graph.add_edge("state_store", "websearch")
         self.graph.add_edge("websearch", "review")
         self.graph.add_edge("response", END)
 
@@ -185,7 +190,9 @@ def process_content(
     """
     summarized_information = (
         state["model"]
-        .invoke(content_prompt, {"configurable": {"thread_id": self.thread_id}})
+        .invoke(
+            content_prompt, {"configurable": {"thread_id": state["thread_id"]}}
+        )
         .content
     )
     return summarized_information
@@ -217,7 +224,7 @@ def main():
         inputs,
         {
             "recursion_limit": 10000,
-            "configurable": {"thread_id": self.thread_id},
+            "configurable": {"thread_id": 42},
         },
     )
 

ursa/prompt_library/hypothesizer_prompts.py

@@ -10,7 +10,7 @@ critic_prompt = dedent("""\
     You are Agent 2, a rigorous Critic who identifies flaws and areas for improvement.
 """)
 
-competitor_prompt = dedent(f"""\
+competitor_prompt = dedent("""\
     You are Agent 3, taking on the role of a direct competitor to Agent 1 in this hypothetical situation.
     Acting as that competitor, and taking into account potential critiques from the critic, provide an honest
     assessment how you might *REALLY* counter the approach of Agent 1.

ursa/util/diff_renderer.py

@@ -1,5 +1,5 @@
-import re
 import difflib
+import re
 from dataclasses import dataclass
 
 from rich.console import Console, ConsoleOptions, RenderResult

ursa/util/memory_logger.py

@@ -1,7 +1,7 @@
 import os
 import shutil
 from pathlib import Path
-from typing import Any, Dict, List, Optional, Sequence
+from typing import Any, Dict, Optional, Sequence
 
 from langchain_chroma import Chroma
 from langchain_core.documents import Document

{ursa_ai-0.2.8.dist-info → ursa_ai-0.2.10.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ursa-ai
-Version: 0.2.8
+Version: 0.2.10
 Summary: Agents for science at LANL
 Author-email: Mike Grosskopf <mikegros@lanl.gov>, Nathan Debardeleben <ndebard@lanl.gov>, Rahul Somasundaram <rsomasundaram@lanl.gov>, Isaac Michaud <imichaud@lanl.gov>, Avanish Mishra <avanish@lanl.gov>, Arthur Lui <alui@lanl.gov>, Russell Bent <rbent@lanl.gov>, Earl Lawrence <earl@lanl.gov>
 License-Expression: BSD-3-Clause

ursa_ai-0.2.10.dist-info/RECORD

@@ -0,0 +1,26 @@
+ursa/agents/__init__.py,sha256=IdwVBLlz7QydHiFSok06McIfEJed6rxrT01hZgpQig0,1008
+ursa/agents/arxiv_agent.py,sha256=6pycpm1hkUNjwXF6lJ4DCF5CxOiSnS5dvpW2Qes3ZD8,14125
+ursa/agents/base.py,sha256=uFhRLVzqhFbTZVA7IePKbUi03ATCXuvga7rzwaHy1B0,1321
+ursa/agents/code_review_agent.py,sha256=aUDq5gT-jdl9Qs-Wewj2oz1d60xov9sN-DOYRfGNTU0,11550
+ursa/agents/execution_agent.py,sha256=-At1EcKRHP9lYQ80jpqTPtQyPuQV-sIpf9J8LlEfWdA,16618
+ursa/agents/hypothesizer_agent.py,sha256=rSLohNQz3xvEcL_DGTFivf9q5BlX1cqlLUcts4GJIjM,23309
+ursa/agents/mp_agent.py,sha256=UyJSheMGHZpWQJL3EgYgPPqArfv6F8sndN05q4CPtyo,6015
+ursa/agents/planning_agent.py,sha256=AKWQJ848RLPiwQGrvDNdN9lBlf3YI5qWmt2hqXnRGj8,5426
+ursa/agents/recall_agent.py,sha256=bQk7ZJtiO5pj89A50OBDzAJ4G2F7ZdsMwmKnp1WWR7g,813
+ursa/agents/websearch_agent.py,sha256=rCv4AWbqe5Us4FmuypM6jptri21nKoNg044ncsu9u3E,8014
+ursa/prompt_library/code_review_prompts.py,sha256=-HuhwW9W_p2LDn44bXLntxLADHCOyl-2KIXxRHto66w,2444
+ursa/prompt_library/execution_prompts.py,sha256=JBBmzVV0605uwFXNv0pxH0fXHqtmOgcDzabjpq3wt2A,2153
+ursa/prompt_library/hypothesizer_prompts.py,sha256=ieupOF5tUy_u8actOjPbK-y5Qkrgw6EYxAfw6RXBebs,762
+ursa/prompt_library/literature_prompts.py,sha256=zhBiN3Q-1Z2hp-hkXXp0T8Ipc-6YUM9gw85DjNu1F6I,421
+ursa/prompt_library/planning_prompts.py,sha256=C8IfVc3ny_5-03bJZop2Yax7wfqS_UIdUGsTZSNQRC0,3534
+ursa/prompt_library/websearch_prompts.py,sha256=n4DJaYn_lIYAVtdy00CCJjT-dLWhn2JNipYqMJAotdY,8846
+ursa/tools/run_command.py,sha256=sQRuHtRyJYWEyL9dpW_Ukc-xQ5vmKKJK1i_6z3uKEfA,690
+ursa/tools/write_code.py,sha256=DtCsUMZegYm0mk-HMPG5Zo3Ba1gbGfnXHsv1NZTdDs8,1220
+ursa/util/diff_renderer.py,sha256=1L1q2qWWb8gLhR532-LgJn2TrqXDx0gUpPVOWD_sqeU,4086
+ursa/util/memory_logger.py,sha256=-4jZkMFXLnABj9x_DMGEUySVPLZaI47HrLgK69Naxw0,5731
+ursa/util/parse.py,sha256=M0cjyQWmjatxX4WbVmDRUiirTLyW-t_Aemlrlrsc5nA,2811
+ursa_ai-0.2.10.dist-info/licenses/LICENSE,sha256=4Vr6_u2zTHIUvYjoOBg9ztDbfpV3hyCFv3mTCS87gYU,1482
+ursa_ai-0.2.10.dist-info/METADATA,sha256=K84_6yoDH8d01VRPhSp42D9z7ODAuJh_7Y58EvXYGAA,6849
+ursa_ai-0.2.10.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+ursa_ai-0.2.10.dist-info/top_level.txt,sha256=OjA1gRYSUAeiXGnpqPC8iOOGfcjFO1IlP848qMnYSdY,5
+ursa_ai-0.2.10.dist-info/RECORD,,

ursa_ai-0.2.8.dist-info/RECORD

@@ -1,26 +0,0 @@
-ursa/agents/__init__.py,sha256=fhuGhYRfOSJ5o2yxjOvdEqJFy2j-VWLDkhiNi3wYNRw,498
-ursa/agents/arxiv_agent.py,sha256=nLZ1huMREr-INSDv1RMmOM8a-DcLBjNBWLSi3oFI5hw,14274
-ursa/agents/base.py,sha256=kGbiGn8qu1eKQSv2Y9YZWEv8ngYsyUMTBkAxk8iD9R0,1334
-ursa/agents/code_review_agent.py,sha256=yVO7nzYI3o0k2HguFw3OSY1IrCy5W8V0YiriYAeviY4,11562
-ursa/agents/execution_agent.py,sha256=Hw7EZem8qYedXjeLi5RDPrPIhqlW1G-vE66MlZ7g1BY,16607
-ursa/agents/hypothesizer_agent.py,sha256=p3bLHyqsiGRwYS4nycYcwnpye2j1umWdaOYspGAFRU0,23309
-ursa/agents/mp_agent.py,sha256=PvCbneJslKSVqWysoGX2_DTYb_JKBWQrYuJvZOVDlw4,6104
-ursa/agents/planning_agent.py,sha256=5KSRk_gDsUrv_6zSxd7CqXhhMCYtnlfNlxSI9tSbqzc,5422
-ursa/agents/recall_agent.py,sha256=UcNRZLbx3j3cHaLEZul4__KzWV4SnUhLTjX9GeoYbHM,823
-ursa/agents/websearch_agent.py,sha256=zDS4IF-WJgsvSmV42HEO582rt3zCh_fJjteh7VpSNe4,7715
-ursa/prompt_library/code_review_prompts.py,sha256=-HuhwW9W_p2LDn44bXLntxLADHCOyl-2KIXxRHto66w,2444
-ursa/prompt_library/execution_prompts.py,sha256=JBBmzVV0605uwFXNv0pxH0fXHqtmOgcDzabjpq3wt2A,2153
-ursa/prompt_library/hypothesizer_prompts.py,sha256=is1SpCbsUtsYsyWOFz3H6M4nCnfyOMPj2p0mOcaEucc,763
-ursa/prompt_library/literature_prompts.py,sha256=zhBiN3Q-1Z2hp-hkXXp0T8Ipc-6YUM9gw85DjNu1F6I,421
-ursa/prompt_library/planning_prompts.py,sha256=C8IfVc3ny_5-03bJZop2Yax7wfqS_UIdUGsTZSNQRC0,3534
-ursa/prompt_library/websearch_prompts.py,sha256=n4DJaYn_lIYAVtdy00CCJjT-dLWhn2JNipYqMJAotdY,8846
-ursa/tools/run_command.py,sha256=sQRuHtRyJYWEyL9dpW_Ukc-xQ5vmKKJK1i_6z3uKEfA,690
-ursa/tools/write_code.py,sha256=DtCsUMZegYm0mk-HMPG5Zo3Ba1gbGfnXHsv1NZTdDs8,1220
-ursa/util/diff_renderer.py,sha256=gHawyUtBLeOq32A25_etDSy-HXAPyZQrnzfYGtHoEIQ,4086
-ursa/util/memory_logger.py,sha256=Qu8JRjqvXvchnVh6s-91te_xnfOAK1fJDyf1DvsRWnI,5737
-ursa/util/parse.py,sha256=M0cjyQWmjatxX4WbVmDRUiirTLyW-t_Aemlrlrsc5nA,2811
-ursa_ai-0.2.8.dist-info/licenses/LICENSE,sha256=4Vr6_u2zTHIUvYjoOBg9ztDbfpV3hyCFv3mTCS87gYU,1482
-ursa_ai-0.2.8.dist-info/METADATA,sha256=AoP6uLIGtPXutp46JzTFKUSv_oh5cXCWQHLtQcCH5eg,6848
-ursa_ai-0.2.8.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-ursa_ai-0.2.8.dist-info/top_level.txt,sha256=OjA1gRYSUAeiXGnpqPC8iOOGfcjFO1IlP848qMnYSdY,5
-ursa_ai-0.2.8.dist-info/RECORD,,