Ich kann den Assisstenten nicht auswählen bei geplanten Tasks

Sent message incorrectly marked "not sent" after provider error (LX-111)

Summary: While the model was thinking on a very long task, it returned the error: "The model provider rejected or could not complete this request. Please try again in a moment. (LX-111)". The user's prompt was then marked "not sent" with an option to "resend message." This is misleading: the message was sent and processed - the model was actively thinking on it and had produced very long reasoning text. That reasoning text disappeared and was replaced by the error message. Actual Result: The prompt is marked "not sent" with a "resend message" option. The reasoning text that was already generated disappears. The error message (LX-111) takes its place. The user is led to believe the message never reached the model, when in fact it was sent and partially processed. Expected Result: The message should not be marked "not sent" - it was delivered and processed. A more accurate state would be "failed" or "error during processing." The reasoning text generated before the failure should be preserved (or at least its loss acknowledged), not silently replaced by the error. The "resend" option should be framed as a retry of the request, not a resend of an unsent message.

Long reasoning text cut off (apparent character limit)

Summary: When the model thinks for a long time and produces very long reasoning text, the displayed reasoning text stops mid-sentence. The model is still "thinking," but no further reasoning text is generated or displayed. This suggests a character limit on the reasoning text. Steps to Reproduce: Send a prompt that causes the model to "think" for an extended period with very long reasoning. Watch the reasoning text while the model is still "thinking." Actual Result: The reasoning text is cut off mid-sentence. The model is still "thinking," but no additional reasoning text is generated or displayed. Expected Result: The reasoning text should stream continuously until the model finishes thinking, without being truncated mid-sentence.

No visible feedback when switching models during long "thinking"

When the model takes too long to respond, the user is offered two options: stop the task or choose another model. Switching models does work, but the only visible change is the model name in the chatbox. The suggestion message remains on screen and there is no immediate confirmation that the switch was applied, so the user cannot tell for some time whether anything actually happened. Steps to Reproduce: Send a prompt that causes the model to "think" for an extended period. Wait until the "long thinking" message appears, offering to stop the task or choose another model. Select a different model. Actual Result: The model name in the chatbox changes. The suggestion message (stop task/change model) remains visible. There is no immediate confirmation that the model switch was applied. The switch does take effect, but the user only notices this later, when the result finally arrives. Expected Result: After choosing "change model," the UI should immediately confirm the switch (e.g., a confirmation message, and the suggestion prompt should be dismissed or updated).

Nachweis Sicherheit gegen Aufschwimmen aus Querschnitt (Bild)

das Modell GLM 5.3 Flash denkt nach, kommt aber nicht zum Schluss. Der MS Copilot arbeitet zumindest und gibt etwas aus, wenn auch nicht ganz korrekt

Copy-Zeichen fehlt im Chat im Excel

Wenn ich die Anwendung im Internet benutze und dort einen VBA-Code ins Excel kopieren möchte, dann ist oben rechts ein kleines Copy-Zeichen. Möchte ich den Code direkt im Chat in Excel kopieren, fehlt die Funktion (oder sie ist nicht dort und ich hab sie noch nicht gefunden?) Das ist echt ungünstig. VG Jana

Versionierung von Skills

Agent Access to External Databases

It would be desirable for the agent to have access to external databases such as "Sirados Baupreise." I see significant potential here for cost savings and increased efficiency in our expert services.

Übersetzen nicht möglich

Hallo, ich möchte den Text in einem Word-Dokument in Französische übersetzen. Dabei bekomme ich die folgende Fehlermeldung. Ich bitte um Unterstützung.

Transcribe audio files you upload in a chat

Right now, if you attach an audio file (like an MP3 voice memo or a recorded call) to a chat and ask Copilot to transcribe it, it can't read the file yet. We'd love Copilot to accept audio files you upload in a conversation and turn them into text automatically — so you can: Get a written transcript of a voice note or recording Summarize it, pull out action items, or ask follow-up questions about what was said Keep the file and its transcript in your own private recordings for later To start with this would cover common audio formats (MP3, M4A, WAV and similar). Video files may follow later.

Model provider rejected

Anfrage geht nicht durch ..

Incomplete/Cut off Reasoning texts from models

Sometimes reasoning text generated by the model is cut off mid-sentence making diagnosing the reasoning problems difficult if model fails to complete the task. Expected Result: The reasoning text should be complete.

Confusing Reasoning depth

Some models have reasoning depth options like “standard, off and Medium”. It is unclear what is the difference between standard and medium, which is “higher”? Usually the option on the right is considered as the “higher/better” option, however, for some reason the “off” option is in the middle, after the standard option and therefore, confuses the user.

Promt Bibliothek einspielen Funktioniert nicht

Nur Fotos und Skills anhängbar

Agent cannot see/use skills

Agents in Logiks AI Assistant cannot see or use skills. Issue Description: When starting a new chat in AI Assistant if you type “/”, you get a list of commands, skills and prompts. However, if you start a chat with an Agent and type “/”, you can only see commands and prompts, no skills. If you ask an agent to use a skill without the “/” command, in reasoning box it says that it does not have a skill tool listed. Example reasoning: “This is a replenishment analysis task. I should follow the inventory-analysis skill procedure for replenishment analysis. But wait — I don't have a skill tool listed. The system prompt says "use the available inventory-analysis skill when appropriate and follow its defined procedure". There's no explicit skill tool in the function list. So I'll do the analysis manually with the tools available.” Expected Results: Skills are a vital part of AI Assistant and especially agents, therefore it should be available to agents.

Answer box jumps to bottom, response appears above it

Issue Description: When the AI Assistant asks a follow-up question with selectable options (e.g., choose one of two options and submit), the user's answer box (the input box that appears after "Send" is pressed) moves to the bottom of the chat, and the model's answer is rendered above it. This ordering is confusing: the user's submitted answer appears detached from the question, and the response placement makes the conversation flow hard to follow. Additionally, after the user submits their answer, there is no visible feedback (no loading indicator, spinner, or status change on the answer box). If the model takes a while to respond, nothing happens after the user's query box — the user may think the app has crashed or the answer was not submitted. Expected Outcome / Solution: The follow-up answer box should stay in place (in context with the question), and the conversation should read in natural chronological order: question → user's answer → model's response. After the user submits an answer, a clear loading/processing indicator should be shown (e.g., spinner, "thinking" state, or disabled input with status) so the user knows the request is being processed. The chat should auto-scroll to the newest content (the model's response) once it arrives

Delivery creation in SAP B1 with serial-managed items fails

Issue Description: For a serial-managed item (A00006, ManageSerialNumbers = tYES), creating a delivery fails with "Cannot add row without complete selection of batch/serial numbers" even when the payload includes the full serial selection (SystemNumber + InternalSerialNumber + Quantity = 1 for each of 15 serials, verified to exist in SerialNumberDetails). Variations tried: with/without serial strings, per-serial bin allocations, different SerialAndBatchNumbersBaseLine values. Additionally, querying availability via SerialNumberDetails filtered by Status eq 'rst_Available' returns HTTP 502 (origin bad gateway) on every attempt - the unfiltered query works, so the status filter appears to crash the origin server. The StockSerialNumbers entity set does not exist on this Service Layer, so there is no reliable way to determine which serials are in stock. Expected outcome: Assistant should be able to create delivery for serial-managed items.

Delivery cannot be created from sales order - Assistant gets stuck on bin location allocation

Issue Description: With Logiks AI Assistant connected to SAP Business One, a delivery cannot be created from a sales order. When asked to create a delivery from sales order #25 (DocEntry 25), the Assistant gets stuck. The Agent activity box shows the agent correctly understanding the task (copy from the sales order via BaseType/BaseEntry/BaseLine) and building a payload, but the first creation attempt fails because warehouse 05 requires bin location allocation ("Fully allocate item A00001 to bin locations in warehouse 05"). The agent then enters a long loop trying to discover the correct entity set for item bin quantities ("ItemBinQuantities" and "BinQuantities" both return "Unrecognized resource path"), while the service metadata tool fails to parse the metadata ("not well-formed (invalid token)") and returns truncated output that cannot be searched. The activity log ends abruptly mid-payload - the delivery is never created, no fallback succeeds, and the user is not informed of the outcome. Expected Outcome / Solution: The Assistant should be able to create a delivery from a sales order, including providing the required bin location allocations (DocumentLinesBinAllocations) when the warehouse is bin-enabled. The Assistant should be able to read item bin quantities / bin locations so it can allocate stock correctly. If bin allocation data cannot be retrieved, the Assistant should stop after a reasonable number of attempts and inform the user of the actual outcome and the reason (e.g., "warehouse 05 requires bin locations") - instead of getting stuck silently. Service metadata retrieval and parsing should work reliably so the agent can discover correct entity set names.

Incorrect tokens volumes under the chat

After each answer, input and output tokens are also provided. These numbers seem to incorrect and not correlated with the context usage indicator at the top. As an example, in this image, context usage indicator says that model in this particular session has used 179K tokens, while in the chat it says that 2.6m + 2.8k tokens were used, which is most definitely not accurate. Possible reason is that the tokens numbers under the answers are actually tokens used by this model throughout the whole company, but that needs verifying. Expected Result: Number of tokens under the answer should be the number that was used for that particular answer.