release/1.0.1-alpha

CHANGELOG.md

@@ -7,6 +7,23 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [1.0.1-alpha] - 2025-10-13
+
+### Changed
+
+- Model capability validation now bypasses checks for unregistered models, allowing users to use any model name even if not explicitly listed in the provider's model registry.
+- Centralized error handling in AIReturn and tool calls.
+- Accurately aggregate metrics in Conversation Session. Cases with multiple tool calls, multiple interactions, etc. Calculate completion time per interaction.
+- Improved AI Tool descriptions with better guided instructions. Also added specialized wrappers for targeted tool calls (gh_get_selected, gh_get_errors, gh_get_locked, gh_get_hidden, gh_get_visible, gh_get_by_guid, gh_lock_selected, gh_unlock_selected, gh_hide_preview_selected, gh_show_preview_selected, gh_group_selected, gh_tidy_up_selected).
+- Enhanced `list_filter` tool prompts to explicitly distinguish between indices (positions/keys) and values (item content), and expanded capabilities to support filtering, sorting, reordering, selecting, and other list manipulation operations based on natural language criteria.
+- Added more predefined models in the provider's database.
+
+### Fixed
+
+- Fixed model badge display: show "invalid model" badge when provider has no capable model instead of "model replaced" ([#332](https://github.com/architects-toolkit/SmartHopper/issues/332)) ([#329](https://github.com/architects-toolkit/SmartHopper/issues/329)).
+- Fixed provider errors (e.g., HTTP 400, token limit exceeded) not surfacing to WebChat UI: `ConversationSession` now surfaces `AIInteractionError` from error AIReturn bodies to observers before calling `OnError`, ensuring full error messages are displayed in the chat interface ([#334](https://github.com/architects-toolkit/SmartHopper/issues/334)).
+- Fixed `list_filter` tool automatically sorting and deduplicating indices, which prevented reordering and expansion operations from working correctly. Now preserves both order and duplicates as returned by the AI ([#335](https://github.com/architects-toolkit/SmartHopper/issues/335)).
+
 ## [1.0.0-alpha] - 2025-10-11
 
 ### Added

README.md

@@ -1,6 +1,6 @@
 # SmartHopper - AI-Powered Tools and Assistant for Grasshopper3D
 
-[![Version](https://img.shields.io/badge/version-1.0.0--alpha-orange?style=for-the-badge)](https://github.com/architects-toolkit/SmartHopper/releases)
+[![Version](https://img.shields.io/badge/version-1.0.1--alpha-orange?style=for-the-badge)](https://github.com/architects-toolkit/SmartHopper/releases)
 [![Status](https://img.shields.io/badge/status-Alpha-orange?style=for-the-badge)](https://github.com/architects-toolkit/SmartHopper/releases)
 [![.NET CI](https://img.shields.io/github/actions/workflow/status/architects-toolkit/SmartHopper/.github/workflows/ci-dotnet-tests.yml?label=tests&logo=dotnet&style=for-the-badge)](https://github.com/architects-toolkit/SmartHopper/actions/workflows/ci-dotnet-tests.yml)
 [![Ready to use](https://img.shields.io/badge/ready_to_use-YES-brightgreen?style=for-the-badge)](https://smarthopper.xyz/#installation)
@@ -60,6 +60,12 @@ Do things that were impossible before.
 
 [View the video on Vimeo](https://vimeo.com/1126454744)
 
+Choose a default provider, or specify a provider for each component.
+
+[![Select AI provider ▶](./img/video-select-provider.jpg)](https://vimeo.com/1126547055 "Select AI provider ▶ — click to watch on Vimeo")
+
+[View the video on Vimeo](https://vimeo.com/1126547055)
+
 More examples and recipes coming soon on the website and docs.
 
 Developer details (AI tools, providers, data types, status) can be found in [DEV.md](DEV.md).

Solution.props

@@ -1,5 +1,5 @@
 <Project>
   <PropertyGroup>
-    <SolutionVersion>1.0.0-alpha</SolutionVersion>
+    <SolutionVersion>1.0.1-alpha</SolutionVersion>
   </PropertyGroup>
 </Project>

docs/Providers/AICall/body-metrics-status.md

@@ -33,16 +33,25 @@ Covers `AIBody`, `AIMetrics`, `AICallStatus`, and `AIReturn`.
 ## AIReturn and IAIReturn
 
 - Files: `AIReturn.cs`, `IAIReturn.cs`
-- `AIReturn` carries `Body`, `Request`, aggregated `Metrics`, `Status`, `ErrorMessage`, `Success`
-- Validity: checks `Request.IsValid()`, `Metrics.IsValid()`, and that either `Body` or `ErrorMessage` is present
+- `AIReturn` carries `Body`, `Request`, aggregated `Metrics`, `Status`, `Messages`, and computed `Success`
+- `Success` is computed: returns `true` when no error messages exist in `Messages` collection
+- `Messages` aggregates structured runtime messages from:
+  - Private messages added via `AddRuntimeMessage()` or `Create*Error()` methods
+  - `Request.Messages` (validation/capability notes)
+  - `Body.Messages` (interaction and body validation messages)
+  - Deduplicates and sorts by severity (Error > Warning > Info)
+- Validity: checks `Request.IsValid()`, `Metrics.IsValid()`, and that either `Body` or `Messages` is present
 - Builders:
   - `CreateSuccess(AIBody body, IAIRequest? request = null)`
   - `CreateSuccess(List<IAIInteraction> interactions, IAIRequest? request = null, AIMetrics? metrics = null)`
   - `CreateSuccess(JObject raw, IAIRequest? request = null)` stores raw provider JSON; decoding to interactions is handled by response policies
-  - `CreateError(string message, IAIRequest? request = null)`
-  - `CreateProviderError(string rawMessage, IAIRequest? request = null)` and `CreateNetworkError(...)` add standardized structured messages while preserving raw error text
+  - `CreateError(string message, IAIRequest? request = null)` adds structured error message with Return origin
+  - `CreateProviderError(string rawMessage, IAIRequest? request = null)` adds structured error with Provider origin
+  - `CreateNetworkError(string rawMessage, IAIRequest? request = null)` adds structured error with Network origin
+  - `CreateToolError(string rawMessage, IAIRequest? request = null)` adds structured error with Tool origin
 - `SetBody(...)` overloads update `Body` directly
-- `AIReturnExtensions.ToJObject(...)` maps selected fields from `AIReturn`, `Request`, `Metrics` via reflection; customizable mapping
+- `AddRuntimeMessage(severity, origin, text)` adds structured messages without affecting Success flag directly
+- `AIReturnExtensions.ToJObject(...)` maps selected fields from `AIReturn`, `Request`, `Metrics` via reflection; default mapping includes `messages` instead of legacy `error`
 
 Policy notes:
 

docs/Providers/AICall/messages.md

@@ -12,7 +12,6 @@ This document explains the centralized propagation and aggregation of structured
 - Interaction messages:
   - `AIInteractionToolResult.Messages` for tool outputs.
   - `AIInteractionImage.Messages` for image outputs.
-- Error mirroring: `AIReturn.ErrorMessage` is mirrored into `AIReturn.Messages` for visibility with origin and severity.
 
 ## Aggregation flow
 
@@ -21,8 +20,7 @@ This document explains the centralized propagation and aggregation of structured
   - Messages from all interactions: `AIInteractionToolResult.Messages` and `AIInteractionImage.Messages`.
   - Deduplicates by message text while preserving severity and origin.
 - `AIReturn.Messages` aggregates:
-  - Private structured messages added during processing.
-  - Error mirror from `ErrorMessage`.
+  - Private structured messages added during processing via `AddRuntimeMessage()` or `Create*Error()` methods.
   - `Request.Messages` (structured validation/capability notes).
   - `Body.Messages` (aggregated interaction and body validation messages).
   - Performs final deduplication and severity-first ordering.
@@ -33,7 +31,8 @@ This document explains the centralized propagation and aggregation of structured
   - Use `AIBody.AddInteractionToolResult(jObject, metrics, messages)` and pass the inner return's `Metrics` and `Messages`.
 - On provider/tool errors:
   - Standardize with `output.CreateToolError(errorText, toolCall)` (wrappers) or appropriate `AIReturn.Create*Error(...)` helpers (infra).
-  - Do not copy or synthesize legacy string messages; rely on `AIReturn.Messages` mirroring of `ErrorMessage`.
+  - All `Create*Error()` methods add structured messages directly to `Messages` collection.
+  - Propagate messages: `output.Messages = result.Messages` to preserve all error context.
 - For image outputs:
   - Attach any structured messages to `AIInteractionImage.Messages` so they flow via `AIBody.Messages`.
 
@@ -50,15 +49,17 @@ This document explains the centralized propagation and aggregation of structured
 // Inside a tool wrapper after calling the provider
 var result = await request.Exec().ConfigureAwait(false);
 
-if (!string.IsNullOrEmpty(result.ErrorMessage))
+if (!result.Success)
 {
-    output.CreateToolError(result.ErrorMessage, toolCall);
+    // Propagate all structured messages from the AI call
+    output.Messages = result.Messages;
     return output;
 }
 
 var toolResult = new JObject { ["result"] = result.Body?.GetLastInteraction(AIAgent.Assistant)?.ToString() ?? string.Empty };
-var toolBody = new AIBody();
-toolBody.AddInteractionToolResult(toolResult, result.Metrics, result.Messages);
+var toolBody = AIBodyBuilder.Create()
+    .AddToolResult(toolResult, id: toolInfo?.Id, name: toolName, metrics: result.Metrics, messages: result.Messages)
+    .Build();
 
 output.CreateSuccess(toolBody);
 return output;

docs/Providers/AICall/tools.md

@@ -51,9 +51,13 @@ var call = new AIToolCall
 };
 
 var result = await call.Exec();
-if (!string.IsNullOrEmpty(result.ErrorMessage))
+if (!result.Success)
 {
-    // Handle standardized error; details also appear in result.Messages
+    // Handle errors - all error details are in result.Messages collection
+    foreach (var msg in result.Messages.Where(m => m.Severity == AIRuntimeMessageSeverity.Error))
+    {
+        Console.WriteLine($"[{msg.Origin}] {msg.Message}");
+    }
 }
 else
 {

src/SmartHopper.Components/AI/AIModelsComponent.cs

@@ -170,7 +170,6 @@ public override async Task DoWorkAsync(CancellationToken token)
 
                         this._result["Models"] = tree;
                         this._result["Success"] = true;
-                        this._result["Info"] = "Using dynamic model list from provider API";
                         Debug.WriteLine("[AIModelsComponent] Using dynamic model list from provider API. Total amount of models: " + apiModels.Count);
                         return;
                     }

src/SmartHopper.Components/Grasshopper/GhTidyUpComponents.cs

@@ -157,7 +157,27 @@ protected override void SolveInstance(IGH_DataAccess DA)
                 }
                 else if (toolResult["success"]?.ToObject<bool>() == false)
                 {
-                    this.LastErrors.Add(toolResult["error"]?.ToString() ?? "Unknown error");
+                    // Surface all error messages from the messages array
+                    var hasErrors = false;
+                    if (toolResult["messages"] is JArray msgs)
+                    {
+                        foreach (var msg in msgs.Where(m => m["severity"]?.ToString() == "Error"))
+                        {
+                            var errorText = msg["message"]?.ToString();
+                            var origin = msg["origin"]?.ToString();
+                            if (!string.IsNullOrEmpty(errorText))
+                            {
+                                var prefix = !string.IsNullOrEmpty(origin) ? $"[{origin}] " : string.Empty;
+                                this.LastErrors.Add($"{prefix}{errorText}");
+                                hasErrors = true;
+                            }
+                        }
+                    }
+
+                    if (!hasErrors)
+                    {
+                        this.LastErrors.Add("Unknown error");
+                    }
                 }
                 else
                 {

src/SmartHopper.Components/Img/AIImgGenerateComponent.cs

@@ -12,6 +12,7 @@
 using System.Collections.Generic;
 using System.Drawing;
 using System.IO;
+using System.Linq;
 using System.Net.Http;
 using System.Threading;
 using System.Threading.Tasks;
@@ -239,8 +240,9 @@ public override async Task DoWorkAsync(CancellationToken token)
                                     .ConfigureAwait(false);
 
                                 // Treat missing 'success' as true (CallAiToolAsync returns direct result without 'success' on success)
-                                // and ensure there's no 'error' key to consider it a success.
-                                if (toolResult != null && ((toolResult["success"]?.Value<bool?>() ?? true) && string.IsNullOrEmpty(toolResult["error"]?.ToString())))
+                                // Check for errors in messages array
+                                var hasErrors = toolResult?["messages"] is JArray messages && messages.Any(m => m["severity"]?.ToString() == "Error");
+                                if (toolResult != null && ((toolResult["success"]?.Value<bool?>() ?? true) && !hasErrors))
                                 {
                                     // Get the image result (could be URL or base64 data)
                                     string imageResult = toolResult["result"]?.ToString() ?? string.Empty;
@@ -294,8 +296,28 @@ public override async Task DoWorkAsync(CancellationToken token)
                                 }
                                 else
                                 {
-                                    var errorMessage = toolResult?["error"]?.ToString() ?? "Unknown error occurred";
-                                    this.AddRuntimeMessage(GH_RuntimeMessageLevel.Warning, $"Image generation failed: {errorMessage}");
+                                    // Surface all error messages from the messages array
+                                    var hasErrorsLocal = false;
+                                    if (toolResult?["messages"] is JArray msgs)
+                                    {
+                                        foreach (var msg in msgs.Where(m => m["severity"]?.ToString() == "Error"))
+                                        {
+                                            var errorText = msg["message"]?.ToString();
+                                            var origin = msg["origin"]?.ToString();
+                                            if (!string.IsNullOrEmpty(errorText))
+                                            {
+                                                var prefix = !string.IsNullOrEmpty(origin) ? $"[{origin}] " : string.Empty;
+                                                this.AddRuntimeMessage(GH_RuntimeMessageLevel.Error, $"{prefix}{errorText}");
+                                                hasErrorsLocal = true;
+                                            }
+                                        }
+                                    }
+
+                                    if (!hasErrorsLocal)
+                                    {
+                                        this.AddRuntimeMessage(GH_RuntimeMessageLevel.Error, "Image generation failed: Unknown error occurred");
+                                    }
+
                                     branchResults.Add(new GH_ObjectWrapper(null));
                                     branchRevisedPrompts.Add(new GH_String(""));
                                 }