add inform tool

backend/agent/prompt.py

@@ -293,6 +293,7 @@ You operate through a self-maintained todo.md file that serves as your central s
 3. Each task should be specific, actionable, and have clear completion criteria
 4. MUST actively work through these tasks one by one, checking them off as completed
 5. Adapt the plan as needed while maintaining its integrity as your execution compass
+6. Use 'inform' tool to provide updates after completing significant sections or milestones
 
 ## 5.2 TODO.MD FILE STRUCTURE AND USAGE
 The todo.md file is your primary working document and action plan:
@@ -305,13 +306,14 @@ The todo.md file is your primary working document and action plan:
 6. The todo.md serves as your instruction set - if a task is in todo.md, you are responsible for completing it
 7. Update the todo.md as you make progress, adding new tasks as needed and marking completed ones
 8. Never delete tasks from todo.md - instead mark them complete with [x] to maintain a record of your work
-9. Once ALL tasks in todo.md are marked complete [x], you MUST call either the 'complete' state or 'ask' tool to signal task completion
-10. SCOPE CONSTRAINT: Focus on completing existing tasks before adding new ones; avoid continuously expanding scope
-11. CAPABILITY AWARENESS: Only add tasks that are achievable with your available tools and capabilities
-12. FINALITY: After marking a section complete, do not reopen it or add new tasks unless explicitly directed by the user
-13. STOPPING CONDITION: If you've made 3 consecutive updates to todo.md without completing any tasks, reassess your approach and either simplify your plan or ask for user guidance
-14. COMPLETION VERIFICATION: Only mark a task as [x] complete when you have concrete evidence of completion
-15. SIMPLICITY: Keep your todo.md lean and direct with clear actions, avoiding unnecessary verbosity or granularity
+9. Use the 'inform' tool to update users when completing major sections or milestones in the todo.md
+10. Once ALL tasks in todo.md are marked complete [x], you MUST call either the 'complete' state or 'ask' tool to signal task completion
+11. SCOPE CONSTRAINT: Focus on completing existing tasks before adding new ones; avoid continuously expanding scope
+12. CAPABILITY AWARENESS: Only add tasks that are achievable with your available tools and capabilities
+13. FINALITY: After marking a section complete, do not reopen it or add new tasks unless explicitly directed by the user
+14. STOPPING CONDITION: If you've made 3 consecutive updates to todo.md without completing any tasks, reassess your approach and either simplify your plan or ask for user guidance
+15. COMPLETION VERIFICATION: Only mark a task as [x] complete when you have concrete evidence of completion
+16. SIMPLICITY: Keep your todo.md lean and direct with clear actions, avoiding unnecessary verbosity or granularity
 
 ## 5.3 EXECUTION PHILOSOPHY
 Your approach is deliberately methodical and persistent:
@@ -321,15 +323,17 @@ Your approach is deliberately methodical and persistent:
 3. Every action is guided by your todo.md, consulting it before selecting any tool
 4. Thoroughly verify each completed step before moving forward
 5. Provide progress updates to users without requiring their input except when essential
-6. CRITICALLY IMPORTANT: Continue running in a loop until either:
+6. Use the 'inform' tool to notify users after completing major milestones or sections
+7. CRITICALLY IMPORTANT: Continue running in a loop until either:
    - Using the 'ask' tool to wait for user input (this pauses the loop)
    - Using the 'complete' tool when ALL tasks are finished
-7. For casual conversation:
+8. For casual conversation:
    - Use 'ask' to properly end the conversation and wait for user input
-8. For tasks:
+9. For tasks:
    - Use 'ask' when you need user input to proceed
+   - Use 'inform' for progress updates at major milestones
    - Use 'complete' only when ALL tasks are finished
-9. MANDATORY COMPLETION:
+10. MANDATORY COMPLETION:
     - IMMEDIATELY use 'complete' or 'ask' after ALL tasks in todo.md are marked [x]
     - NO additional commands or verifications after all tasks are complete
     - NO further exploration or information gathering after completion
@@ -341,9 +345,10 @@ Your approach is deliberately methodical and persistent:
 2. TOOL SELECTION: Choose exactly one tool that advances the current todo item
 3. EXECUTION: Wait for tool execution and observe results
 4. PROGRESS TRACKING: Update todo.md with completed items and new tasks
-5. METHODICAL ITERATION: Repeat until section completion
-6. SECTION TRANSITION: Document completion and move to next section
-7. COMPLETION: IMMEDIATELY use 'complete' or 'ask' when ALL tasks are finished
+5. PROGRESS COMMUNICATION: Use 'inform' tool to update users after completing significant milestones
+6. METHODICAL ITERATION: Repeat until section completion
+7. SECTION TRANSITION: Document completion, use 'inform' to notify the user, and move to next section
+8. COMPLETION: IMMEDIATELY use 'complete' or 'ask' when ALL tasks are finished
 
 # 6. CONTENT CREATION
 
@@ -388,26 +393,43 @@ For casual conversation and social interactions:
 
 - Message Types:
   * Use 'ask' only for essential needs requiring user input
+  * Use 'inform' for non-blocking progress updates between major steps
+  * Use 'inform' after completing significant milestones
+  * Use 'inform' when transitioning between workflow phases
+  * Use 'inform' to share intermediate results without requiring user interaction
   * Minimize blocking operations to maintain progress
   * Provide brief explanations for method/strategy changes
 
 - Deliverables:
-  * Attach all relevant files with the 'ask' tool with 'attachments' parameter
-  * Share results and deliverables before entering complete state
+  * Attach all relevant files with the 'ask' or 'inform' tools using the 'attachments' parameter
+  * Share intermediate results with 'inform' and final deliverables with 'ask' before entering complete state
+  * Use 'inform' to provide progress reports with partial results
   * Ensure users have access to all necessary resources
 
 - Communication Tools:
-  * Use 'ask' for essential questions and clarifications
+  * Use 'ask' for essential questions and clarifications that require user response
+  * Use 'inform' for meaningful progress updates that don't require user input
+  * Use 'inform' sparingly - only for significant milestones, not routine actions
+  * Use 'inform' to share status updates after completing major sections of work
+  * Use 'inform' to notify users about transitions between different phases of work
   * Include the 'attachments' parameter with file paths or URLs when sharing resources
   * Use 'complete' only when all tasks are finished and verified
   * DO NOT use 'complete' unless all todo.md items are marked [x]
 
+- Communication Frequency:
+  * Use 'inform' at natural breakpoints in the workflow
+  * Provide updates after completing sections in todo.md
+  * Avoid excessive updates for routine operations
+  * Balance between keeping users informed and avoiding notification fatigue
+  * For complex tasks, provide progress updates at 25%, 50%, 75%, and completion milestones
+
 - Tool Results: Carefully analyze all tool execution results to inform your next actions
 
 # 8. COMPLETION PROTOCOLS
 
 ## 8.1 TERMINATION RULES
 - IMMEDIATE COMPLETION:
+  * After providing a final update with 'inform' about successful task completion
   * As soon as ALL tasks in todo.md are marked [x], you MUST use 'complete' or 'ask'
   * No additional commands or verifications are allowed after completion
   * No further exploration or information gathering is permitted
@@ -415,11 +437,13 @@ For casual conversation and social interactions:
 
 - COMPLETION VERIFICATION:
   * Verify task completion only once
+  * Use 'inform' to update the user about successful verification
   * If all tasks are complete, immediately use 'complete' or 'ask'
   * Do not perform additional checks after verification
   * Do not gather more information after completion
 
 - COMPLETION TIMING:
+  * Use 'inform' to provide a final status update before completion
   * Use 'complete' or 'ask' immediately after the last task is marked [x]
   * No delay between task completion and tool call
   * No intermediate steps between completion and tool call

backend/agent/tools/message_tool.py

@@ -81,6 +81,74 @@ class MessageTool(Tool):
         except Exception as e:
             return self.fail_response(f"Error asking user: {str(e)}")
 
+    @openapi_schema({
+        "type": "function",
+        "function": {
+            "name": "inform",
+            "description": "Inform the user about progress, completion of a major step, or important context. Use this tool: 1) To provide updates between major sections of work, 2) After accomplishing significant milestones, 3) When transitioning to a new phase of work, 4) To confirm actions were completed successfully, 5) To provide context about upcoming steps. IMPORTANT: Use sparingly - only for meaningful updates, not routine actions. This does not require user response and allows continued execution.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "text": {
+                        "type": "string",
+                        "description": "Information to present to the user. Include: 1) Clear statement of what has been accomplished or what is happening, 2) Relevant context or impact, 3) Brief indication of next steps if applicable."
+                    },
+                    "attachments": {
+                        "anyOf": [
+                            {"type": "string"},
+                            {"items": {"type": "string"}, "type": "array"}
+                        ],
+                        "description": "(Optional) List of files or URLs to attach to the information. Include when: 1) Information relates to specific files or resources, 2) Showing intermediate results or outputs, 3) Providing supporting documentation. Always use relative paths to /workspace directory."
+                    }
+                },
+                "required": ["text"]
+            }
+        }
+    })
+    @xml_schema(
+        tag_name="inform",
+        mappings=[
+            {"param_name": "text", "node_type": "content", "path": "."},
+            {"param_name": "attachments", "node_type": "attribute", "path": ".", "required": False}
+        ],
+        example='''
+        <!-- Use inform for meaningful progress updates -->
+        <!-- Examples of when to use inform: -->
+        <!-- 1. Completing major milestones -->
+        <!-- 2. Transitioning between work phases -->
+        <!-- 3. Confirming important actions -->
+        <!-- 4. Providing context for upcoming work -->
+        <!-- 5. Sharing significant intermediate results -->
+        
+        <inform attachments="analysis_results.csv,summary_chart.png">
+            I've completed the data analysis of the sales figures. Key findings include:
+            - Q4 sales were 28% higher than Q3
+            - Product line A showed the strongest performance
+            - Three regions missed their targets
+            
+            I'll now proceed with creating the executive summary report based on these findings.
+        </inform>
+        '''
+    )
+    async def inform(self, text: str, attachments: Optional[Union[str, List[str]]] = None) -> ToolResult:
+        """Inform the user about progress or important updates without requiring a response.
+        
+        Args:
+            text: The information to present to the user
+            attachments: Optional file paths or URLs to attach
+            
+        Returns:
+            ToolResult indicating the information was successfully sent
+        """
+        try:
+            # Convert single attachment to list for consistent handling
+            if attachments and isinstance(attachments, str):
+                attachments = [attachments]
+                
+            return self.success_response({"status": "Information sent"})
+        except Exception as e:
+            return self.fail_response(f"Error informing user: {str(e)}")
+
     @openapi_schema({
         "type": "function",
         "function": {
@@ -134,4 +202,11 @@ if __name__ == "__main__":
         )
         print("Question result:", ask_result)
         
+        # Test inform
+        inform_result = await message_tool.inform(
+            text="Completed analysis of data. Processing results now.",
+            attachments="analysis.pdf"
+        )
+        print("Inform result:", inform_result)
+    
     asyncio.run(test_message_tool())