bioguider 0.2.3__py3-none-any.whl → 0.2.4__py3-none-any.whl

bioguider/agents/common_agent_2step.py

@@ -113,6 +113,95 @@ class CommonAgentTwoSteps(CommonAgent):
             logger.error(str(e))
             raise e
         processed_res = None
+        if post_process is not None:
+            try:
+                processed_res = post_process(res, **kwargs)
+            except RetryException as e:
+                logger.error(str(e))
+                self.exceptions = [e] if self.exceptions is None else self.exceptions + [e]
+                raise e
+            except Exception as e:
+                logger.error(str(e))
+                raise e
+        return res, processed_res, self.token_usage, reasoning_process
+    
+FINAL_STEP_SYSTEM_PROMPTS = ChatPromptTemplate.from_template("""
+---
+
+You will be given a response generated by a LLM, which includes a **step-by-step reasoning process** followed by a clearly marked **final answer**.
+
+### **Your Task:**
+
+Extract and return only the content of the **final answer**.
+
+---
+
+### **Important Instructions:**
+1. Your task is to **extract only the final answer** from the provided reasoning process.
+   **Do not** make any judgments, interpretations, or modifications to the content.
+
+### **Input:**
+
+{llm_response}
+
+---
+""")
+
+class CommonAgentTwoChainSteps(CommonAgentTwoSteps):
+    def __init__(self, llm):
+        super().__init__(llm)
+
+    def _invoke_agent(self, system_prompt, instruction_prompt, schema, post_process = None, **kwargs):
+        # Initialize the callback handler
+        callback_handler = OpenAICallbackHandler()
+        processed_system_prompt = system_prompt.replace("{", "(").replace("}", ")")
+        cot_prompt = self._build_prompt_for_cot_step(
+            system_prompt=processed_system_prompt, 
+            instruction_prompt=instruction_prompt
+        )
+
+        try:
+            # First, use llm to do CoT
+            msgs = cot_prompt.invoke(input={}).to_messages()
+            
+            cot_res = self.llm.generate(messages=[msgs])
+            if cot_res is None or cot_res.llm_output is None:
+                raise Exception("llm generate invalid output")
+            reasoning_process = cot_res.generations[0][0].text
+            token_usage: Any = cot_res.llm_output.get("token_usage")
+            cot_tokens = {
+                "total_tokens": token_usage.get("total_tokens", 0),
+                "prompt_tokens": token_usage.get("prompt_tokens", 0),
+                "completion_tokens": token_usage.get("completion_tokens", 0),
+            }
+            self._incre_token_usage(cot_tokens)
+        except Exception as e:
+            logger.error(str(e))
+            raise e
+                
+        try:
+            # Then use the reasoning process to do the structured output
+            processed_reasoning_process = reasoning_process.replace("{", "{{").replace("}", "}}")
+            final_msg = FINAL_STEP_SYSTEM_PROMPTS.format(
+                llm_response=processed_reasoning_process,
+            )
+            msgs = [(
+                "human",
+                final_msg,
+            )]
+            final_prompt = ChatPromptTemplate.from_messages(msgs)
+            agent = final_prompt | self.llm.with_structured_output(schema)
+            res = agent.invoke(
+                input={},
+                config={
+                    "callbacks": [callback_handler],
+                },
+            )
+            self._incre_token_usage(callback_handler)
+        except Exception as e:
+            logger.error(str(e))
+            raise e
+        processed_res = None
         if post_process is not None:
             try:
                 processed_res = post_process(res, **kwargs)

bioguider/agents/dockergeneration_observe_step.py

@@ -6,7 +6,7 @@ from pydantic import BaseModel, Field
 from bioguider.utils.constants import DEFAULT_TOKEN_USAGE
 from bioguider.agents.agent_utils import run_command, read_file
 from bioguider.agents.dockergeneration_task_utils import DockerGenerationWorkflowState
-from bioguider.agents.common_agent_2step import CommonAgentTwoSteps
+from bioguider.agents.common_agent_2step import CommonAgentTwoChainSteps, CommonAgentTwoSteps
 from bioguider.agents.peo_common_step import PEOCommonStep
 
 DOCKERGENERATION_OBSERVE_SYSTEM_PROMPT = """You are an expert in software containerization and reproducibility engineering.
@@ -93,7 +93,7 @@ class DockerGenerationObserveStep(PEOCommonStep):
             if code != 0:
                 error_msg = DockerGenerationObserveStep._extract_error_message(error)
                 system_prompt = self._build_system_prompt(state, error_msg, "N/A")
-                agent = CommonAgentTwoSteps(llm=self.llm)
+                agent = CommonAgentTwoChainSteps(llm=self.llm)
                 res, _, token_usage, reasoning = agent.go(
                     system_prompt=system_prompt,
                     instruction_prompt="Now, let's begin observing.",
@@ -125,7 +125,7 @@ class DockerGenerationObserveStep(PEOCommonStep):
                     "docker build successfully.",
                     error,
                 )
-                agent = CommonAgentTwoSteps(llm=self.llm)
+                agent = CommonAgentTwoChainSteps(llm=self.llm)
                 res, _, token_usage, reasoning = agent.go(
                     system_prompt=system_prompt,
                     instruction_prompt="Now, let's begin observing.",

bioguider/agents/dockergeneration_plan_step.py

@@ -12,7 +12,7 @@ from bioguider.agents.agent_utils import (
     PlanAgentResultJsonSchema,
 )
 from bioguider.agents.peo_common_step import PEOCommonStep
-from bioguider.agents.common_agent_2step import CommonAgentTwoSteps
+from bioguider.agents.common_agent_2step import CommonAgentTwoChainSteps, CommonAgentTwoSteps
 from bioguider.agents.dockergeneration_task_utils import (
     DockerGenerationWorkflowState, 
     prepare_provided_files_string,
@@ -140,7 +140,7 @@ class DockerGenerationPlanStep(PEOCommonStep):
 
     def _execute_directly(self, state: DockerGenerationWorkflowState):
         system_prompt = self._prepare_system_prompt(state)
-        agent = CommonAgentTwoSteps(llm=self.llm)
+        agent = CommonAgentTwoChainSteps(llm=self.llm)
         res, _, token_usage, reasoning = agent.go(
             system_prompt=system_prompt,
             instruction_prompt="Now, let's begin to make a plan",

bioguider/agents/evaluation_task.py

@@ -2,14 +2,15 @@
 import os
 from pathlib import Path
 import logging
-from typing import Callable
+from typing import Callable, Optional
 from abc import ABC, abstractmethod
 from langchain.prompts import ChatPromptTemplate
 from langchain_openai.chat_models.base import BaseChatOpenAI
+from pydantic import BaseModel, Field
 
 from bioguider.agents.agent_utils import read_file
-from bioguider.utils.constants import ProjectMetadata
-from .common_agent_2step import CommonAgentTwoSteps
+from bioguider.utils.constants import DEFAULT_TOKEN_USAGE, ProjectMetadata
+from .common_agent_2step import CommonAgentTwoSteps, CommonAgentTwoChainSteps
 from .common_agent import CommonConversation
 from ..utils.pyphen_utils import PyphenReadability
 from ..utils.gitignore_checker import GitignoreChecker
@@ -19,12 +20,19 @@ logger = logging.getLogger(__name__)
 EVALUATION_README_SYSTEM_PROMPT = """
 You are an expert in evaluating the quality of README files in software repositories. Your task is to analyze the provided README file and generate a comprehensive quality report.
 
-### **README Quality Report**
+---
 
-For each criterion below, provide a brief assessment followed by specific, actionable comments for improvement.
+### **Step 1:  Identify README type
+
+First, determine whether the provided README is a **project-level README** (typically at the root of a repository) or a **folder-level README** (typically inside subdirectories).
 
 ---
 
+### **Project-level README Evaluation**
+
+If the README is a **project-level** file, evaluate it using the following criteria.
+For each criterion below, provide a brief assessment followed by specific, actionable comments for improvement.
+
 **1. Project Clarity & Purpose**
  * **Assessment**: [Your evaluation of whether the project's purpose is clear.]
  * **Improvement Suggestions**:
@@ -62,11 +70,53 @@ For each criterion below, provide a brief assessment followed by specific, actio
  * **SMOG Index**: `{smog_index}` (Estimates the years of education needed to understand the text).
  * **Assessment**: Based on these scores, evaluate the overall readability and technical complexity of the language used.
 
-**7. Overall Quality Summary**
+**Final Answer**
+ * Project-Level README: Yes / No
  * Provide a final, overall assessment of the README file's quality, summarizing the key strengths and areas for improvement.
 
 ---
 
+### **Folder-Level README Evaluation**
+
+If the README is a **folder-level** file, use the following criteria instead.
+
+For each criterion below, provide a brief assessment followed by specific, actionable comments for improvement.
+
+**1. Folder Description**
+ * **Assessment**: [Your evaluation of whether it Provides a clear **description** of what the folder contains (e.g., modules, scripts, data).]
+ * **Improvement Suggestions**:
+    * **Original text:** [Quote a specific line/section from the README.]
+    * **Improving comments:** [Provide your suggestions to improve clarity.]
+
+**2. Folder Purpose**
+ * **Assessment**: [Your evaluation of whether it explains the **purpose** or **role** of the components inside this subfolder.]
+ * **Improvement Suggestions**:
+    * **Original text:** [Quote text related to purpose.]
+    * **Improving comments:** [Provide your suggestions.]
+
+**3. Usage**
+ * **Assessment**: [Your evaluation of whether it includes **usage instructions** specific to this folder (e.g., commands, import paths, input/output files).]
+ * **Improvement Suggestions**:
+    * **Original text:** [Quote text related to usage.]
+    * **Improving comments:** [Provide your suggestions.]
+
+**4. Readability Analysis**
+ * **Flesch Reading Ease**: `{flesch_reading_ease}` (A higher score is better, with 60-70 being easily understood by most adults).
+ * **Flesch-Kincaid Grade Level**: `{flesch_kincaid_grade}` (Represents the US school-grade level needed to understand the text).
+ * **Gunning Fog Index**: `{gunning_fog_index}` (A score above 12 is generally considered too hard for most people).
+ * **SMOG Index**: `{smog_index}` (Estimates the years of education needed to understand the text).
+ * **Assessment**: Based on these scores, evaluate the overall readability and technical complexity of the language used.
+
+**Final Answer**
+ * Project-Level README: Yes / No
+ * Provide a final, overall assessment of the README file's quality, summarizing the key strengths and areas for improvement.
+---
+
+### **README path:**
+{readme_path}
+
+---
+
 ### **README Content:**
 {readme_content}
 """
@@ -100,11 +150,11 @@ class EvaluationTask(ABC):
             token_usage=token_usage,
         )
 
-    def evaluate(self, files: list[str] | None = None):
+    def evaluate(self, files: list[str] | None = None) -> dict:
         self._enter_evaluation()
-        evaluation, token_usage = self._evaluate(files)
+        evaluations, token_usage = self._evaluate(files)
         self._leave_evaluation(token_usage)
-        return evaluation
+        return evaluations
     
     def _enter_evaluation(self):
         self.print_step(step_name=self.evaluation_name)
@@ -113,9 +163,13 @@ class EvaluationTask(ABC):
         self.print_step(token_usage=token_usage)
 
     @abstractmethod
-    def _evaluate(self, files: list[str]):
+    def _evaluate(self, files: list[str]) -> tuple[dict, dict]:
         pass
 
+class EvaluationREADMEResult(BaseModel):
+    project_level: Optional[bool]=Field(description="a boolean value specifying if the README file is **project-level** README. TRUE: project-level, FALSE, folder-level")
+    overall_assessment: Optional[str]=Field(description="an overall assessment")
+
 class EvaluationREADMETask(EvaluationTask):
     def __init__(
         self, 
@@ -128,7 +182,7 @@ class EvaluationREADMETask(EvaluationTask):
         super().__init__(llm, repo_path, gitignore_path, meta_data, step_callback)
         self.evaluation_name = "README Evaluation"
             
-    def _evaluate(self, files: list[str]):
+    def _evaluate(self, files: list[str]) -> tuple[dict, dict]:
         readme_files = files
         if readme_files is None or len(readme_files) == 0:
             return None
@@ -140,6 +194,8 @@ class EvaluationREADMETask(EvaluationTask):
             if readme_content is None:
                 logger.error(f"Error in reading file {readme_file}")
                 continue
+            if len(readme_content.strip()) == 0:
+                readme_content = "empty file"
 
             readability = PyphenReadability()
             flesch_reading_ease, flesch_kincaid_grade, gunning_fog_index, smog_index, \
@@ -148,19 +204,28 @@ class EvaluationREADMETask(EvaluationTask):
                 EVALUATION_README_SYSTEM_PROMPT
             ).format(
                 readme_content=readme_content,
+                readme_path=readme_file,
                 flesch_reading_ease=flesch_reading_ease,
                 flesch_kincaid_grade=flesch_kincaid_grade,
                 gunning_fog_index=gunning_fog_index,
                 smog_index=smog_index,
             )
-            conversation = CommonConversation(llm=self.llm)
-            response, token_usage = conversation.generate(
+            # conversation = CommonConversation(llm=self.llm)
+            agent = CommonAgentTwoChainSteps(llm=self.llm)
+            response, _, token_usage, reasoning_process = agent.go(
                 system_prompt=system_prompt,
-                instruction_prompt="Before arriving at the conclusion, clearly explain your reasoning step by step. Now, let's begin the evaluation."
+                instruction_prompt="Before arriving at the conclusion, clearly explain your reasoning step by step. Now, let's begin the evaluation.",
+                schema=EvaluationREADMEResult,
             )
             self.print_step(step_output=f"README: {readme_file}")
-            self.print_step(step_output=response)
-            readme_evaluations[readme_file] = response
+            self.print_step(step_output=reasoning_process)
+            readme_evaluations[readme_file] = {
+                "evaluation": {
+                    "project_level": response.project_level,
+                    "overall_assessment": response.overall_assessment,
+                }, 
+                "reasoning_process": reasoning_process
+            }
         return readme_evaluations, token_usage
         
 EVALUATION_TUTORIAL_SYSTEM_PROMPT="""
@@ -233,9 +298,9 @@ class EvaluationTutorialTask(EvaluationTask):
         super().__init__(llm, repo_path, gitignore_path, meta_data, step_callback)
         self.evaluation_name = "Tutorial Evaluation"
 
-    def _evaluate(self, files: list[str]):
+    def _evaluate(self, files: list[str]) -> tuple[dict, dict]:
         if len(files) == 0:
-            return None
+            return {}, {**DEFAULT_TOKEN_USAGE}
         
         evaluations = {}
         for file in files:

bioguider/agents/identification_observe_step.py

@@ -2,7 +2,7 @@
 from langchain.prompts import ChatPromptTemplate
 
 from bioguider.agents.agent_utils import ObservationResult
-from bioguider.agents.common_agent_2step import CommonAgentTwoSteps
+from bioguider.agents.common_agent_2step import CommonAgentTwoSteps, CommonAgentTwoChainSteps
 from bioguider.agents.identification_task_utils import IdentificationWorkflowState
 from bioguider.agents.peo_common_step import PEOWorkflowState, PEOCommonStep
 
@@ -70,7 +70,7 @@ class IdentificationObserveStep(PEOCommonStep):
 
     def _execute_directly(self, state: IdentificationWorkflowState):
         system_prompt = self._prepare_system_prompt(state)
-        agent = CommonAgentTwoSteps(llm=self.llm)
+        agent = CommonAgentTwoChainSteps(llm=self.llm)
         res, _, token_usage, reasoning_process = agent.go(
             system_prompt=system_prompt,
             instruction_prompt="Now, let's begin.",

bioguider/agents/identification_plan_step.py

@@ -5,7 +5,7 @@ from langchain.tools import BaseTool
 from pydantic import BaseModel, Field
 
 from bioguider.agents.agent_utils import get_tool_names_and_descriptions
-from bioguider.agents.common_agent_2step import CommonAgentTwoSteps
+from bioguider.agents.common_agent_2step import CommonAgentTwoChainSteps, CommonAgentTwoSteps
 from bioguider.agents.identification_task_utils import IdentificationWorkflowState
 from bioguider.agents.peo_common_step import PEOCommonStep
 
@@ -113,7 +113,7 @@ class IdentificationPlanStep(PEOCommonStep):
 
     def _execute_directly(self, state: IdentificationWorkflowState):
         system_prompt = self._prepare_system_prompt(state)
-        agent = CommonAgentTwoSteps(llm=self.llm)
+        agent = CommonAgentTwoChainSteps(llm=self.llm)
         res, _, token_usage, reasoning_process = agent.go(
             system_prompt=system_prompt,
             instruction_prompt="Now, let's begin.",

bioguider/managers/evaluation_manager.py

@@ -57,7 +57,7 @@ class EvaluationManager:
         )
         return self.project_metadata
     
-    def evaluate_readme(self):
+    def evaluate_readme(self) -> tuple[any, list[str]]:
         task = EvaluationREADMETask(
             llm=self.llm,
             repo_path=self.rag.repo_dir,
@@ -67,7 +67,7 @@ class EvaluationManager:
         )
         readme_files = self._find_readme_files()
         results = task.evaluate(readme_files)
-        return results
+        return results, readme_files
     
     def evaluate_tutorial(self):
         task = CollectionTask(
@@ -97,7 +97,7 @@ class EvaluationManager:
         repo_path = self.rag.repo_dir
         gitignore_path = Path(repo_path, ".gitignore")
         gitignore_checker = GitignoreChecker(
-            directory=self.repo_path, gitignore_path=gitignore_path
+            directory=repo_path, gitignore_path=gitignore_path
         )
         found_readme_files = gitignore_checker.check_files_and_folders(
             check_file_cb=lambda root_dir, relative_path: Path(relative_path).name.lower() in possible_readme_files,

bioguider/rag/data_pipeline.py

@@ -518,6 +518,21 @@ class DatabaseManager:
         self.reset_database()
         self._create_repo(repo_url_or_path, access_token)
         return self.prepare_db_index()
+    
+    def _extract_repo_name_from_url(self, repo_url_or_path: str, repo_type: str) -> str:
+        # Extract owner and repo name to create unique identifier
+        url_parts = repo_url_or_path.rstrip('/').split('/')
+
+        if repo_type in ["github", "gitlab", "bitbucket"] and len(url_parts) >= 5:
+            # GitHub URL format: https://github.com/owner/repo
+            # GitLab URL format: https://gitlab.com/owner/repo or https://gitlab.com/group/subgroup/repo
+            # Bitbucket URL format: https://bitbucket.org/owner/repo
+            owner = url_parts[-2]
+            repo = url_parts[-1].replace(".git", "")
+            repo_name = f"{owner}_{repo}"
+        else:
+            repo_name = url_parts[-1].replace(".git", "")
+        return repo_name
 
     def reset_database(self):
         """
@@ -545,19 +560,18 @@ class DatabaseManager:
             root_path = retrieve_data_root_path()
 
             os.makedirs(root_path, exist_ok=True)
+            repo_type = "unknown"
             # url
             if repo_url_or_path.startswith("https://") or repo_url_or_path.startswith("http://"):
                 # Extract repo name based on the URL format
                 if "github.com" in repo_url_or_path:
                     # GitHub URL format: https://github.com/owner/repo
-                    repo_name = repo_url_or_path.split("/")[-1].replace(".git", "")
+                    repo_type = "github"
                 elif "gitlab.com" in repo_url_or_path:
                     # GitLab URL format: https://gitlab.com/owner/repo or https://gitlab.com/group/subgroup/repo
                     # Use the last part of the URL as the repo name
-                    repo_name = repo_url_or_path.split("/")[-1].replace(".git", "")
-                else:
-                    # Generic handling for other Git URLs
-                    repo_name = repo_url_or_path.split("/")[-1].replace(".git", "")
+                    repo_type = "gitlab"
+                repo_name = self._extract_repo_name_from_url(repo_url_or_path, repo_type)
 
                 save_repo_dir = os.path.join(root_path, "repos", repo_name)
 

bioguider/utils/file_utils.py

@@ -117,9 +117,11 @@ def parse_repo_url(url: str) -> tuple[str | None, str | None]:
     except Exception:
         return None, None
 
-def retrieve_data_root_path():
+def retrieve_data_root_path() -> Path:
     data_folder = os.environ.get("DATA_FOLDER", "./data")
-    return data_folder
+    root_folder = Path(data_folder, ".adalflow")
+    return root_folder.absolute()
+    
 
 
 

{bioguider-0.2.3.dist-info → bioguider-0.2.4.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: bioguider
-Version: 0.2.3
+Version: 0.2.4
 Summary: An AI-Powered package to help biomedical developers to generate clear documentation
 License: MIT
 Author: Cankun Wang

{bioguider-0.2.3.dist-info → bioguider-0.2.4.dist-info}/RECORD

@@ -9,17 +9,17 @@ bioguider/agents/collection_plan_step.py,sha256=mx-_5Y3pqKDPBaMMyFElKlpq1GWN7g03
 bioguider/agents/collection_task.py,sha256=9jqasC5PVZvnPzQtLGAJiYvO1M2RGWH7tqWO6KbEnl8,6663
 bioguider/agents/collection_task_utils.py,sha256=WRzzpMV6r8aY0FlX_zroHbLDXyrmvS48OSiBr_fIq2Q,3677
 bioguider/agents/common_agent.py,sha256=eGs8m8bjO0dmW6lDIen7DQNdWdHD7j8Udf3XhL1k6vI,5242
-bioguider/agents/common_agent_2step.py,sha256=wIg9m9-8tq0AYrG2quuI1ZYg82Q54MHtTQLG4pqN8y4,4596
+bioguider/agents/common_agent_2step.py,sha256=IJ5SxqsK26oj8W3U4wnGtbJxHRrHEznaGCYFBXKUHn4,7916
 bioguider/agents/common_step.py,sha256=GdOCbmj1pwh4etg-futVFYVDQuoUG89DnIrw-B6QbzM,2594
 bioguider/agents/dockergeneration_execute_step.py,sha256=F92jDlkc6KjAvTkX7q1FsCYP8J15SCaNgmwh3YPqfDo,6500
-bioguider/agents/dockergeneration_observe_step.py,sha256=gVCsr0EiKCgYMRQvev_LWBf78yHzLxIltPVIgRmdS3o,6062
-bioguider/agents/dockergeneration_plan_step.py,sha256=v0X95Lzj5kjfqcOch6IOMcuOfsRBkgJjsYDlxMpblLI,7204
+bioguider/agents/dockergeneration_observe_step.py,sha256=93PO_Y4YyUShVTKRt0nErcjb-xYTcwcZCj7TgniS9t4,6098
+bioguider/agents/dockergeneration_plan_step.py,sha256=SB8tQM9PkIKsD2o1DFD7bedcxz6r6hSy8n_EVK60Fz0,7235
 bioguider/agents/dockergeneration_task.py,sha256=ezsweVHJsFpOyOI6rYMt1DZ3PE19dcq4J3Lm-d0IA8M,6220
 bioguider/agents/dockergeneration_task_utils.py,sha256=v7emqrJlVW-A5ZdLmPSdiaMSKCR8uzy9UYzx_1cgzyo,9041
-bioguider/agents/evaluation_task.py,sha256=iEa47N1iL-MgLKNZfMzRjjMcg2rEqSis4FeXaEGYLJQ,10928
+bioguider/agents/evaluation_task.py,sha256=RRcbw4qTUPElZuiBAMj7heGsDV3t93l_IrioudMTGpc,14316
 bioguider/agents/identification_execute_step.py,sha256=_9x_KITVtNJjQlgjfq-LazoUl46Tctaj_W_AVxz5q-w,5488
-bioguider/agents/identification_observe_step.py,sha256=B5E61luYteyKOaMZsgjxI-xuNijPfAjNpNBWL5DyTHc,3636
-bioguider/agents/identification_plan_step.py,sha256=p-a3vc0qhifJCM0cerFVAtYuYSvqImzqCsQofFaQRrs,5193
+bioguider/agents/identification_observe_step.py,sha256=OENwf9XyOSIHvJMp7eoyQOYGjjtPnPT2S29xf1rCATk,3667
+bioguider/agents/identification_plan_step.py,sha256=p0BKziXdB4ph4D_T9FU5bH8CbHD5Gv0YuszMds_xh-Y,5224
 bioguider/agents/identification_task.py,sha256=-4bnw39OSkhu8RkDVQPRpmu0g8K79dji3HXTybch5As,8329
 bioguider/agents/identification_task_utils.py,sha256=5gevknha9hJiiQN5L7Yp9-pyhAlbR_j31aGRK5j0D_w,522
 bioguider/agents/peo_common_step.py,sha256=iw2c1h7X11WJzSE2tSRg0UAoXH0QOlQDxW9CCzSVMOY,2677
@@ -28,20 +28,20 @@ bioguider/agents/python_ast_repl_tool.py,sha256=o7-4P1h8jS8ikhGSA4CI_OWQ2a0Eg5tE
 bioguider/agents/rag_collection_task.py,sha256=r_jPAMjQcC7dIydKxX77UuMqjJ3MiVKswNZ-yNw7yx8,5199
 bioguider/conversation.py,sha256=DIvk_d7pz_guuORByK1eaaF09FAK-8shcNTrbSUHz9Y,1779
 bioguider/database/summarized_file_db.py,sha256=9Iyin5BChlFjDJlFamTwmzkYZvO3BzFu7M8AbikyPXw,4441
-bioguider/managers/evaluation_manager.py,sha256=EtVSTP75mtlvXDcXKMedsqlwUZEkxaD9MBwHJQfpmr4,3920
+bioguider/managers/evaluation_manager.py,sha256=h-I-ot41pvXsy1Q8vXfJAS_Znb4-68vnwrdIC-OBXeA,3954
 bioguider/rag/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 bioguider/rag/config.py,sha256=5g4IqTzgyfZfax9Af9CTkXShgItPOt4_9TEMSekCPik,4602
-bioguider/rag/data_pipeline.py,sha256=rG3iSxWlIZzQ5ml3E2Zo7UJEnlVeeneAm_kSCsngUhw,27126
+bioguider/rag/data_pipeline.py,sha256=OXnsqETVytHBMXHerg9gACtNhwpWSODYWvzxVDTP_So,27767
 bioguider/rag/embedder.py,sha256=jofR8hOj3Aj2IyBQ9y6FeAc84tgq5agbIfCGyFxYpJ8,650
 bioguider/rag/rag.py,sha256=2G2b7JIDsjrR74lnkIFyEuMPF14kn6B-WhphZkUxd3c,4481
 bioguider/settings.py,sha256=BD_iz9aYarxmWUl0XaKl4-D4oTXMhFzljsXLNn2phis,3143
 bioguider/utils/constants.py,sha256=_xMAhwE3py2RR0pIimnb2qfucXdnTj4ZNeKGACouh2w,932
 bioguider/utils/default.gitignore,sha256=XjPdyO2KV8z8iyuqluaNR_70tBQftMpyKL8HboVNyeI,1605
-bioguider/utils/file_utils.py,sha256=R68ayY1flc2Mbl7yMkp_39AUzw5_8iPF5Ow3YVQaZdY,3610
+bioguider/utils/file_utils.py,sha256=9VfAHsz1UkFPtzAmvWZvPl1TMaKIYNjNlLgsfB8tNjg,3683
 bioguider/utils/gitignore_checker.py,sha256=pOYUwsS9D5014LxcZb0cj3s2CAYaD2uF_pYJpaNKcho,6532
 bioguider/utils/pyphen_utils.py,sha256=cdZc3qphkvMDeL5NiZ8Xou13M_uVNP7ifJ-FwxO-0BE,2680
 bioguider/utils/utils.py,sha256=YP3HXgU_rvYDWkEcTzWGiYZw-mlfVrqGhUGSc0_4Pms,900
-bioguider-0.2.3.dist-info/LICENSE,sha256=qzkvZcKwwA5DuSuhXMOm2LcO6BdEr4V7jwFZVL2-jL4,1065
-bioguider-0.2.3.dist-info/METADATA,sha256=_Vaa_hK-a2Y9bI07jzhC8UW16TLjMlwLjqXVfwOSJzQ,1823
-bioguider-0.2.3.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
-bioguider-0.2.3.dist-info/RECORD,,
+bioguider-0.2.4.dist-info/LICENSE,sha256=qzkvZcKwwA5DuSuhXMOm2LcO6BdEr4V7jwFZVL2-jL4,1065
+bioguider-0.2.4.dist-info/METADATA,sha256=VD_LpYetLIYKw2y-JH7_OZZT8D5oIbmzSCx8t5kgyJY,1823
+bioguider-0.2.4.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
+bioguider-0.2.4.dist-info/RECORD,,