Enhance action execution prompts and improve chat client configuration

FlowVision/lib/Classes/ai/LMStudioActioner.cs

@@ -83,8 +83,16 @@ public async Task<string> ExecuteAction(string actionPrompt)
                 // Add system message to actioner history
                 actionerHistory.Add(new ChatMessage(ChatRole.System, toolConfig.ActionerSystemPrompt + toolDescriptions));
 
-                // Add action prompt to actioner history
-                actionerHistory.Add(new ChatMessage(ChatRole.User, actionPrompt));
+                // Add action prompt with explicit instruction to EXECUTE
+                string enhancedPrompt = $@"{actionPrompt}
+
+IMPORTANT REMINDER:
+1. DO NOT just observe and describe - you must EXECUTE the action!
+2. After CaptureWholeScreen(), you MUST continue to actually click/type/interact
+3. Follow ALL steps: Observe → Plan → EXECUTE → Verify
+4. Do not stop until you've performed the actual action requested";
+
+                actionerHistory.Add(new ChatMessage(ChatRole.User, enhancedPrompt));
 
                 // Verify LM Studio is enabled and configured
                 if (!lmStudioConfig.Enabled)
@@ -162,9 +170,21 @@ public async Task<string> ExecuteAction(string actionPrompt)
                 };
 
                 // Build chat client with function invocation if enabled
-                actionerChat = toolConfig.AutoInvokeKernelFunctions 
-                    ? new ChatClientBuilder(baseChatClient).UseFunctionInvocation().Build()
-                    : baseChatClient;
+                if (toolConfig.AutoInvokeKernelFunctions)
+                {
+                    // Use function invocation with default behavior
+                    // The middleware will automatically loop and call tools
+                    actionerChat = new ChatClientBuilder(baseChatClient)
+                        .UseFunctionInvocation()
+                        .Build();
+
+                    PluginLogger.LogInfo("LMStudioActioner", "ExecuteAction", 
+                        "Function invocation enabled - will auto-invoke tools");
+                }
+                else
+                {
+                    actionerChat = baseChatClient;
+                }
 
                 // Update loading message to show we're now processing the response
                 PluginLogger.StopLoadingIndicator();

docs/Prompt-Engineering-Guide.md

@@ -0,0 +1,64 @@
+---
+layout: default
+title: Prompt Engineering Guide
+---
+
+# Prompt Engineering Guide for Recursive Control
+
+Master the art of communicating with your AI agents to achieve precise, efficient automation.
+
+## Why Prompt Engineering Matters
+Recursive Control's multi-agent system (Hermes, Daedalus, and Talos) interprets natural language, but structured prompts yield better results:
+- **Clarity reduces errors**: Ambiguous requests lead to wrong assumptions.
+- **Structure aids planning**: Well-organized prompts help the Planner agent create better steps.
+- **Context improves accuracy**: Provide relevant details for better execution.
+
+## Core Principles
+1. **Be Specific**: Include exact details like app names, file paths, or expected outcomes.
+2. **Break It Down**: For complex tasks, suggest steps or use multi-stage prompts.
+3. **Provide Context**: Mention current state, like open windows or recent actions.
+4. **Use Verification**: Ask the agent to confirm steps or describe what it sees.
+5. **Handle Errors Gracefully**: Include fallback instructions.
+
+## Prompt Patterns
+### Basic Command
+**Template**: "Perform [action] in [location/app] with [details]."
+
+**Example**: "Open Chrome and navigate to github.com/flowdevs-io/Recursive-Control."
+
+### Multi-Step Workflow
+**Template**: "Do the following steps: 1. [Step 1] 2. [Step 2] ... Verify each step."
+
+**Example**: "Create a report: 1. Open Excel. 2. Add headers: Date, Task, Status. 3. Fill with today's data. 4. Save as 'daily-report.xlsx' in Documents."
+
+### Vision-Assisted
+**Template**: "Capture the screen, describe what you see, then [action based on description]."
+
+**Example**: "Take a screenshot of the current window, identify the search bar, and type 'AI tools' into it."
+
+### Conditional Logic
+**Template**: "If [condition], do [action A]; else do [action B]."
+
+**Example**: "If Chrome is open, navigate to YouTube; else open Chrome first then go to YouTube."
+
+## Advanced Techniques
+### Chain of Thought
+Encourage reasoning: "Think step-by-step: First, check if the app is open. If not, open it. Then..."
+
+### Role Playing
+Assign roles: "As a efficient automation expert, optimize this workflow: [task]."
+
+### Few-Shot Examples
+Provide samples: "Like how you opened Notepad last time, now open Paint and draw a square."
+
+## Common Pitfalls
+- **Too Vague**: "Do something with files" → Agents might guess wrong.
+- **Overly Complex**: Break long prompts into multiple interactions.
+- **Assuming State**: Always verify: "Focus on the foreground window and describe it first."
+
+## Tips for Power Users
+- Use the UI's chat history to build context across prompts.
+- Combine plugins explicitly: "Use Playwright to automate browser, then CMD to process downloads."
+- Test incrementally: Start with simple tasks and build up.
+
+Master these patterns to turn Recursive Control into your ultimate productivity copilot!

docs/Workflow-Builder-Tutorial.md

@@ -0,0 +1,104 @@
+---
+layout: default
+title: Workflow Builder Tutorial
+---
+
+# Workflow Builder Tutorial: Gamified Learning Adventures
+
+Level up your Recursive Control skills by building interactive workflows! This tutorial turns learning into a game: complete "quests" by crafting prompt chains that automate real tasks. Each quest includes objectives, step-by-step guidance, and verification challenges.
+
+Think of this as a simulator – test your prompts here before running them in the app. Earn "badges" by successfully completing each workflow (self-assessed via expected outcomes).
+
+## Quest 1: GitHub Issue Automator (Beginner Level)
+**Objective**: Automate creating a GitHub issue in your repo using browser automation and keyboard inputs.
+
+**Badge**: Issue Master
+
+**Step-by-Step Prompt Chain**:
+1. **Launch the Browser**: "Open Chrome and navigate to github.com/login."
+   - *Expected*: Browser opens to GitHub login page. (Verify: Describe the screen to confirm.)
+
+2. **Login**: "Focus on the username field and type 'yourusername', then tab to password and type 'yourpassword', then press enter."
+   - *Tip*: Use KeyboardPlugin for targeted input. (Challenge: Add verification – "If login fails, notify me.")
+
+3. **Navigate to Repo**: "Go to github.com/yourusername/your-repo/issues."
+   - *Expected*: Issues page loads.
+
+4. **Create Issue**: "Click the 'New issue' button, type 'Bug: App crashes on load' in title, add description 'Steps to reproduce: 1. Open app. 2. Click button.', then submit."
+   - *Challenge*: Use ScreenCapture to identify the button's bounding box before clicking.
+
+**Full Chain Prompt** (Copy-Paste Ready):
+```
+Perform these steps to create a GitHub issue:
+1. Open Chrome and go to github.com/login.
+2. Log in with username 'yourusername' and password 'yourpassword'.
+3. Navigate to github.com/yourusername/your-repo/issues.
+4. Click 'New issue', fill title 'Bug: App crashes', description 'Steps: Open app, click button', and submit.
+Verify each step with a screenshot description.
+```
+
+**Verification Quest**: Run this in Recursive Control. Did it create the issue? If not, refine the prompt (e.g., handle 2FA).
+
+## Quest 2: Daily Report Generator (Intermediate Level)
+**Objective**: Automate opening Excel, filling data, and saving a report.
+
+**Badge**: Report Wizard
+
+**Step-by-Step Prompt Chain**:
+1. **Open App**: "Launch Excel and create a new spreadsheet."
+2. **Add Headers**: "Type 'Date' in A1, 'Task' in B1, 'Status' in C1."
+3. **Fill Data**: "In A2 type today's date, B2 'Implement feature X', C2 'Completed'."
+4. **Save**: "Save as 'daily-report.xlsx' in Documents."
+
+**Full Chain Prompt**:
+```
+Build a daily report in Excel:
+1. Open Excel, new file.
+2. Headers: A1=Date, B1=Task, C1=Status.
+3. Data: A2=today's date, B2=Review code, C2=Done.
+4. Save to Documents as daily-report.xlsx.
+Confirm each cell after typing.
+```
+
+**Challenge**: Add conditional logic – "If file exists, append instead of overwrite."
+
+## Quest 3: File Organizer Bot (Advanced Level)
+**Objective**: Scan Downloads folder, organize files by type using CMD/PowerShell.
+
+**Badge**: Organization Overlord
+
+**Step-by-Step Prompt Chain**:
+1. **Scan Folder**: "Use CMD to list files in Downloads."
+2. **Create Folders**: "Make directories: Images, Documents, Others."
+3. **Move Files**: "Move .jpg to Images, .pdf to Documents, others to Others."
+4. **Verify**: "List contents of each new folder."
+
+**Full Chain Prompt**:
+```
+Organize Downloads:
+1. CMD: dir %USERPROFILE%\Downloads
+2. Create folders: mkdir Images Documents Others
+3. Move: move *.jpg Images, move *.pdf Documents, move *.* Others
+4. Verify: dir Images, dir Documents, dir Others
+Handle errors if folders exist.
+```
+
+**Epic Challenge**: Integrate vision – "Screenshot Downloads folder and describe file icons before moving."
+
+## Level Up Tips
+- **Gamification Hack**: Track your success rate. Aim for 100% on 5 quests to "unlock" custom workflow creation.
+- **Combine Quests**: Chain them, e.g., "Generate report, then create GitHub issue about it."
+- **Debug Mode**: If a step fails, add "Describe the screen and suggest fixes" to your prompts.
+
+## Go Interactive: Jupyter Notebook Simulator
+For hands-on practice, download our [Interactive Workflow Builder Notebook](tutorials/Workflow-Builder-Notebook.ipynb). It includes widgets to build and simulate prompt chains without running the full app!
+
+**How to Use**:
+1. Install Jupyter: `pip install notebook`
+2. Download the .ipynb file.
+3. Run `jupyter notebook` and open the file.
+4. Interact with widgets to test prompts in real-time.
+
+Completed all quests? Share your custom workflows on Discord!
+
+Back to [Getting Started](Getting-Started.html)