Release v4.5.102

Author: MervinPraison

docker/Dockerfile.chat

@@ -16,7 +16,7 @@ RUN mkdir -p /root/.praison
 # Install Python packages (using latest versions)
 RUN pip install --no-cache-dir \
     praisonai_tools \
-    "praisonai>=4.5.101" \
+    "praisonai>=4.5.102" \
     "praisonai[chat]" \
     "embedchain[github,youtube]"
 

docker/Dockerfile.dev

@@ -20,7 +20,7 @@ RUN mkdir -p /root/.praison
 # Install Python packages (using latest versions)
 RUN pip install --no-cache-dir \
     praisonai_tools \
-    "praisonai>=4.5.101" \
+    "praisonai>=4.5.102" \
     "praisonai[ui]" \
     "praisonai[chat]" \
     "praisonai[realtime]" \

docker/Dockerfile.ui

@@ -16,7 +16,7 @@ RUN mkdir -p /root/.praison
 # Install Python packages (using latest versions)
 RUN pip install --no-cache-dir \
     praisonai_tools \
-    "praisonai>=4.5.101" \
+    "praisonai>=4.5.102" \
     "praisonai[ui]" \
     "praisonai[crewai]"
 

src/praisonai-agents/praisonaiagents/agents/agents.py

@@ -665,16 +665,22 @@ async def aexecute_task(self, task_id):
         if executor_agent and executor_agent.tools:
             tools.extend(executor_agent.tools)
 
-        task_prompt = f"""
-You need to do the following task: {task.description}.
-Expected Output: {task.expected_output}.
-"""
+        # Substitute variables in task description if provided
+        task_description = task.description
+        if getattr(task, 'variables', None):
+            for key, value in task.variables.items():
+                task_description = task_description.replace(f"{{{{{key}}}}}", str(value))
+
+        # Build context first to include in task prompt
+        context_text = ""
         if task.context:
             context_results = []  # Use list to avoid duplicates
             for context_item in task.context:
                 # Use the centralized helper function
                 context_str = process_task_context(context_item, self.verbose, self.user_id)
-                context_results.append(context_str)
+                # Only add non-empty context strings
+                if context_str and context_str.strip():
+                    context_results.append(context_str)
 
             # Join unique context results with proper formatting
             unique_contexts = list(dict.fromkeys(context_results))  # Remove duplicates
@@ -683,15 +689,26 @@ async def aexecute_task(self, task_id):
                 for i, ctx in enumerate(unique_contexts):
                     logger.debug(f"Context {i+1}: {ctx[:100]}...")
             context_separator = '\n\n'
-            task_prompt += f"""
-Context:
+            context_text = context_separator.join(unique_contexts)
+        
+        # Build task prompt - only use "User Input/Topic" format if there's actual content
+        if context_text and context_text.strip():
+            task_prompt = f"""
+User Input/Topic: {context_text}
 
-{context_separator.join(unique_contexts)}
-"""
-        task_prompt += "Please provide only the final result of your work. Do not add any conversation or extra explanation."
+Task: {task_description}
+Expected Output: {task.expected_output}
+
+IMPORTANT: Your response must be about the user's input/topic above. Incorporate it into your task.
+Please provide only the final result of your work. Do not add any conversation or extra explanation."""
+        else:
+            task_prompt = f"""
+You need to do the following task: {task_description}.
+Expected Output: {task.expected_output}.
+Please provide only the final result of your work. Do not add any conversation or extra explanation."""
 
         if self.verbose >= 2:
-            logger.info(f"Executing task {task_id}: {task.description} using {executor_agent.display_name}")
+            logger.info(f"Executing task {task_id}: {task_description} using {executor_agent.display_name}")
         logger.debug(f"Starting execution of task {task_id} with prompt:\n{task_prompt}")
 
         if task.images:
@@ -989,16 +1006,16 @@ def execute_task(self, task_id):
             for key, value in task.variables.items():
                 task_description = task_description.replace(f"{{{{{key}}}}}", str(value))
 
-        task_prompt = f"""
-You need to do the following task: {task_description}.
-Expected Output: {task.expected_output}.
-"""
+        # Build context first to include in task prompt
+        context_text = ""
         if task.context:
             context_results = []  # Use list to avoid duplicates
             for context_item in task.context:
                 # Use the centralized helper function
                 context_str = process_task_context(context_item, self.verbose, self.user_id)
-                context_results.append(context_str)
+                # Only add non-empty context strings
+                if context_str and context_str.strip():
+                    context_results.append(context_str)
 
             # Join unique context results with proper formatting
             unique_contexts = list(dict.fromkeys(context_results))  # Remove duplicates
@@ -1007,11 +1024,21 @@ def execute_task(self, task_id):
                 for i, ctx in enumerate(unique_contexts):
                     logger.debug(f"Context {i+1}: {ctx[:100]}...")
             context_separator = '\n\n'
-            task_prompt += f"""
-Context:
+            context_text = context_separator.join(unique_contexts)
+        
+        # Build task prompt - only use "User Input/Topic" format if there's actual content
+        if context_text and context_text.strip():
+            task_prompt = f"""
+User Input/Topic: {context_text}
+
+Task: {task_description}
+Expected Output: {task.expected_output}
 
-{context_separator.join(unique_contexts)}
-"""
+IMPORTANT: Your response must be about the user's input/topic above. Incorporate it into your task."""
+        else:
+            task_prompt = f"""
+You need to do the following task: {task_description}.
+Expected Output: {task.expected_output}."""
 
         # Add memory context if available
         if task.memory:
@@ -1025,7 +1052,7 @@ def execute_task(self, task_id):
             except Exception as e:
                 logger.error(f"Error getting memory context: {e}")
 
-        task_prompt += "Please provide only the final result of your work. Do not add any conversation or extra explanation."
+        task_prompt += "\nPlease provide only the final result of your work. Do not add any conversation or extra explanation."
 
         if self.verbose >= 2:
             logger.info(f"Executing task {task_id}: {task.description} using {executor_agent.display_name}")

src/praisonai-agents/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "praisonaiagents"
-version = "1.5.101"
+version = "1.5.102"
 description = "Praison AI agents for completing complex tasks with Self Reflection Agents"
 readme = "README.md"
 requires-python = ">=3.10"

src/praisonai-agents/uv.lock

@@ -3,8 +3,8 @@ class Praisonai < Formula
 
     desc "AI tools for various AI applications"
     homepage "https://github.com/MervinPraison/PraisonAI"
-    url "https://github.com/MervinPraison/PraisonAI/archive/refs/tags/v4.5.101.tar.gz"
-    sha256 `curl -sL https://github.com/MervinPraison/PraisonAI/archive/refs/tags/v4.5.101.tar.gz | shasum -a 256`.split.first
+    url "https://github.com/MervinPraison/PraisonAI/archive/refs/tags/v4.5.102.tar.gz"
+    sha256 `curl -sL https://github.com/MervinPraison/PraisonAI/archive/refs/tags/v4.5.102.tar.gz | shasum -a 256`.split.first
     license "MIT"
 
     depends_on "python@3.11"

src/praisonai/praisonai.rb

@@ -57,7 +57,7 @@ def create_dockerfile(self):
             file.write("FROM python:3.11-slim\n")
             file.write("WORKDIR /app\n")
             file.write("COPY . .\n")
-            file.write("RUN pip install flask praisonai==4.5.101 gunicorn markdown\n")
+            file.write("RUN pip install flask praisonai==4.5.102 gunicorn markdown\n")
             file.write("EXPOSE 8080\n")
             file.write('CMD ["gunicorn", "-b", "0.0.0.0:8080", "api:app"]\n')
 

src/praisonai/praisonai/deploy.py

@@ -48,11 +48,11 @@ class PraisonAIAgentComponent(Component):
             info="Name for identification and logging.",
             value="Agent",
         ),
-        IntInput(
-            name="order",
-            display_name="Order",
-            info="Execution order in sequential workflows (1 = first, 2 = second, etc.).",
-            value=1,
+        HandleInput(
+            name="previous_agent",
+            display_name="Previous Agent",
+            info="Connect from previous agent to define execution order (Agent 1 → Agent 2 → Agent 3).",
+            input_types=["Agent"],
         ),
         MultilineInput(
             name="role",
@@ -396,8 +396,8 @@ def build_agent(self) -> Any:
         # Build agent
         agent = agent_class(**kwargs)
 
-        # Store order for sequential execution sorting
-        agent._langflow_order = self.order or 1
+        # Store previous agent reference for chaining
+        agent._langflow_previous = self.previous_agent
 
         self.status = f"Agent '{self.agent_name}' created"
         return agent

src/praisonai/praisonai/flow/components/PraisonAI/praisonai_agent.py

@@ -187,6 +187,60 @@ def _import_agents(self):
         else:
             return AgentTeam
 
+    def _sort_agents_by_chain(self, agents: list) -> list:
+        """Sort agents by their chain order (previous_agent connections).
+        
+        Agents are connected: Agent1 → Agent2 → Agent3
+        The first agent has no previous_agent, subsequent agents point to their predecessor.
+        """
+        if len(agents) <= 1:
+            return agents
+        
+        # Find agents that have a previous_agent set
+        has_previous = set()
+        
+        for agent in agents:
+            prev = getattr(agent, '_langflow_previous', None)
+            if prev is not None:
+                has_previous.add(id(agent))
+        
+        # Find starting agents (those with no previous)
+        starting_agents = [a for a in agents if id(a) not in has_previous or getattr(a, '_langflow_previous', None) is None]
+        
+        if not starting_agents:
+            # No clear start, return as-is
+            return agents
+        
+        # Build chain from each starting agent
+        sorted_agents = []
+        visited = set()
+        
+        # Start with agents that have no previous
+        for start in starting_agents:
+            if id(start) in visited:
+                continue
+            sorted_agents.append(start)
+            visited.add(id(start))
+        
+        # Now add agents that follow (have previous_agent set)
+        # Keep iterating until all agents are placed
+        remaining = [a for a in agents if id(a) not in visited]
+        while remaining:
+            added = False
+            for agent in remaining[:]:
+                prev = getattr(agent, '_langflow_previous', None)
+                if prev is not None and id(prev) in visited:
+                    sorted_agents.append(agent)
+                    visited.add(id(agent))
+                    remaining.remove(agent)
+                    added = True
+            if not added:
+                # No progress, add remaining as-is
+                sorted_agents.extend(remaining)
+                break
+        
+        return sorted_agents
+
     def build_agents(self) -> Any:
         """Build and return the PraisonAI Agents instance."""
         agents_class = self._import_agents()
@@ -198,8 +252,8 @@ def build_agents(self) -> Any:
             msg = "At least one agent is required."
             raise ValueError(msg)
 
-        # Sort agents by their order (set in Agent component)
-        agents = sorted(agents, key=lambda a: getattr(a, '_langflow_order', 1))
+        # Sort agents by chain order (previous_agent connections)
+        agents = self._sort_agents_by_chain(agents)
 
         # Build kwargs - tasks auto-generated from agents
         kwargs = {