Wish I could ask Cowork about itself without it actually doing anything
Sharing this one to see if it lands for others using Cowork for governance, design, or any kind of multi-step work. I've started running Cowork sessions for things like Purview labeling design, Entra cleanup planning, and DLP policy scoping. The output is useful, but honestly the most valuable thing in those sessions is the path — what got considered, what got rejected, what trade-offs surfaced. Days or weeks later I want to come back and ask "why did we land here?" without Cowork picking the session back up and changing things. What I'd love to see is a read-only side Chat against a Cowork session. Full read access to the transcript, files, decisions, and artifacts — but zero write or action permissions. Just Q&A. There's also a credit angle that matters to me. If I'm just asking follow-up questions about what was already built, that's a research/read task — I don't want to spend Cowork session credits on it. Today I work around it by launching a regular Chat and pointing it at whatever documentation or artifacts I exported out of the Cowork session. It works, but the Chat doesn't have the full context — it only sees what I exported, not the actual session transcript, intermediate steps, or the paths Cowork considered and rejected. A built-in read-only Chat against the session itself would give me Chat-level efficiency with full Cowork context. The use cases for IT infrastructure architecture and governance work line up cleanly: Research follow-ups ("why X over Y?") Decision review ("what trade-offs were discussed?") Handoff and onboarding ("walk me through what was done") Audit and governance ("explain this outcome to a stakeholder") Cowork would essentially become a durable, queryable record of how complex work was done — not just what came out the other end. For anyone treating Cowork as a system of record for design and decisions, that's a big unlock. Is anyone else thinking about Cowork this way? And for those who have hit this — how are you handling the "explain what Cowork did" problem today? https://feedbackportal.microsoft.com/feedback/idea/2e7b6245-fa6c-f111-9b47-6045bdbd0989
co-pilot and other AI platforms are horrible
I spent more than three hours trying to solve a problem sharing calendars. Co-pilot spun me around in circles. I understand that Microsoft fired many programmers and depend more on co-pilot. This is a huge mistake. after all of the hours, I gave up and now I will have to pay a human to solve my computer problem. I sincerely hope that someone in Microsoft, who has a connection to decision making, will read this a take it seriously. I have used Microsoft hardware and software for more than 40 years, all the way back to Anytime, Anywhere Learning as a trainer and presenter. Could a human who is not interested in corporate profit pay attention to this huge problem.Anyone else hitting the "this Chat just became a Cowork job" wall?
Now that Cowork is GA, I've been using it a lot more in my day-to-day as an IT Infrastructure Architect, and one thing keeps tripping me up. I'll start a quick Copilot Chat for what feels like a one-shot question (scope a policy, sanity-check a PowerShell snippet, sketch a config), and three or four turns in I realize the work is actually multi-step. Exactly what Cowork was built for. But there's no bridge between the two. So, I'm stuck with two bad options: restart in Cowork and lose the context I just built up in Chat, or copy/paste fragments across and hope I didn't drop something important. The other thing that catches me is the credit cost. Chat is a lighter-weight place to explore, and that's where most of my early turns belong. Without a clean bridge, the temptation is to start everything in Cowork "just in case it gets complex," which burns session credits on work that often turned out to be simple. Or I restart Cowork mid-conversation and effectively re-run the same prompts twice. A one-click handoff lets me stay in Chat as long as it makes sense and only graduates to Cowork when the work actually warrants it. What I'd love to see is a simple "Open in Cowork" action in Chat that brings the conversation, referenced files, and any artifacts already produced into a fresh Cowork session, with my latest prompt as the starting task. Chat stays at the low-friction entry point, and Cowork picks up cleanly when the work grows up. Curious whether others are running into the same thing, and if anyone has found a workflow that handles this cleanly today. Always interested in how the broader community is solving for it. I created a User Feedback item for this if anyone wants to upvote it: https://feedbackportal.microsoft.com/feedback/idea/448ca034-c36c-f111-9b47-6045bdbd0989
Subject: Feature Request – Enable Copilot to Retrieve Full Article Content from External URLs
Subject: Feature Request – Enable Copilot to Retrieve Full Article Content from External URLs Dear Copilot Support Team, I am a frequent user of Microsoft Copilot and greatly appreciate its capabilities. Currently, Copilot is unable to retrieve article content directly from external websites such as note.com and exblog.jp when a URL is provided. Other AI services (ChatGPT, Gemini, Grok) can read full article content simply by providing a URL, without requiring the user to manually paste the entire text. For long-form articles, manually copying and pasting the full content is extremely time‑consuming and impractical. Therefore, I would like to request the following improvements: • Enable Copilot to fetch and read full article content directly from URLs • Add support for major blogging platforms such as note.com and exblog.jp • Allow long articles to be analyzed without requiring manual text input I believe this enhancement would significantly improve usability and bring Copilot in line with other leading AI tools. Thank you very much for your consideration.Copilot apps cannot communicate across platforms
I use multiple platforms including Windows, iPad and iPhone. I assumed that the best approach would be to use the dedicated app for each device but I soon discovered that is a mistake. Chat's done on one platform are not visible on the other platforms. I asked Copilot about this and it conceded that was the case and that the recommendation is to use the Web based Copilot rather than the apps. it is a shame to see a company the size of Microsoft put out apps called the same on multiple platforms but those can't pass data to the other platforms. This is a huge productivity issue. It is obvious that there are different teams working on the apps for the different platforms and they are not communicating. This is a real shame and is a pretty big miss for a billion dollar company. Makes me question their long term AI strategy and how competitive they will be.Prompt Lab: Three Critical Bugs That Make It Unusable as a Prompt Management Tool
Product: Microsoft 365 Copilot — Prompt Lab (accessed via ... button in Copilot Chat) Date: June 1, 2026 Environment: M365 Copilot, Web (Edge), Work IQ mode Bug #1: Empty State Crash Steps to reproduce: Open Prompt Lab via the ... button in Copilot Chat Ensure "Your saved prompts" contains zero prompts (either as a new user or by deleting all saved prompts) Observe the result Expected behavior: An empty state placeholder (e.g., "You haven't saved any prompts yet.") Actual behavior: The entire Prompt Lab panel throws a red error banner: "Something went wrong. Please close the dialog and try again later." The "Your saved prompts" category does not render at all. The panel only shows Microsoft's preset categories (Prompt topics, Agent prompts). Why this matters: An empty container is a valid state — it is literally every new user's initial state. A UI component should never crash because a list has zero items. This is a null/empty array handling failure that should have been caught by basic QA. Bug #2: Search Does Not Index User's Saved Prompts Steps to reproduce: Save a custom prompt with a distinctive title (e.g., "AI每日新闻") and body containing the keyword "AI" Open Prompt Lab Use the search box at the bottom to search for "AI" Expected behavior: Search results include the user's saved prompt alongside Microsoft's preset prompts. Actual behavior: Only Microsoft preset prompts matching "AI" are returned (e.g., "Stay on top of AI," "Prompt Compliance"). The user's own saved prompt — whose title and body both contain "AI" — does not appear in the results. Why this matters: The search box creates a false expectation that it searches all prompts. In reality, it only indexes Microsoft's template library. This means as a user accumulates more saved prompts, the only way to find one is manual scrolling. A search function that excludes user-created content is fundamentally broken by design. Bug #3: Saved Prompts Lost During Migration Context: The standalone Copilot Lab / Prompt Gallery app was retired on July 15, 2025, and its functionality was merged into the built-in Prompt Lab within Copilot Chat. What happened: All previously saved prompts from the old Copilot Lab app are gone. They do not appear in the new Prompt Lab's "Your saved prompts" section. There was no migration notice, no export tool for end users, and no recovery path. Why this matters: Users invested time curating and refining their prompt libraries. Silently dropping that data during a platform migration — without warning, backup, or migration tooling — is a breach of user trust. Summary These three issues compound into a single conclusion: Prompt Lab is currently non-functional as a prompt management tool. Capability Status Reliable storage ❌ Data lost during migration Empty state handling ❌ Crashes when empty Search / retrieval ❌ Does not index user content Displaying Microsoft templates ✅ Works The only feature that works correctly is showcasing Microsoft's own preset prompts. For a tool whose entire purpose is to help users save, organize, and reuse their own prompts, this is an unacceptable state of quality. I'd strongly recommend the team prioritize: (1) proper null-state handling, (2) including user prompts in the search index, and (3) investigating whether migrated prompt data can be recovered from Substrate.
Copilot OneDrive and Teams Integration useless for PDFs of scanned documents
Hi, I have this ubiquitous Copilot button everywhere in OneDrive, Teams, SharePoint, you name it. But Everytime I use the suggested „Summarise this file“ button of a PDF document that contains scanned pages Copilot complains that the images are very low resolution and nothing meaningful can be extracted. But when I let Copilot analyse the exact same file in the pure Copilot view everything works fine. After some lengthy conversations with Copilot it admits, that it only access the complete file when the Copilot view is used. In all other views using Copilot only works on the preview images that are created of the file when uploaded. That explains why the images to analyse are of a too low resolution to do OCR and why on some files only the first page gets processed. Why is Copilot integrated in that way? I expect Copilot to always work on the original file for any request I do. Especially when Copilot is promoted everywhere and I’m constantly nagged to get a summary or FAQ of a selected file. The way it currently works would be better removed or -better- just redirect with the selected file into Copilot view. CheersMicrosoft Cowork (Frontier) Scheduled Runs Can't Access Custom Plugins/MCP Tools
Hi everyone, I'm currently building a ServiceNow incident handling automation using Microsoft Cowork (available through the Microsoft Frontier program) and have run into an issue that I haven't been able to resolve. Solution Architecture I have built: A custom FastAPI MCP backend hosted on Azure Web App MCP tools for: Fetching new ServiceNow incidents (state = new) Retrieving relevant ServiceNow Knowledge Base articles based on incident context Updating ServiceNow comments and work notes Performing Microsoft Entra ID automation (e.g., user profile updates requested through incidents) Custom plugins and skills deployed through the Microsoft 365 Admin Center What Works When I start the Cowork agent manually and provide instructions such as processing new incidents: It successfully discovers my custom tools. It fetches incidents from ServiceNow. It retrieves KB articles. It updates work notes/comments. It performs Entra ID actions when required. It processes incidents sequentially according to the workflow. So the overall integration appears to be working correctly. The Problem The issue occurs when I configure the Cowork agent to run on a scheduled basis (every hour). Instead of executing the workflow, the agent reports messages like: and after retries: From the Workspace panel, I can see the scheduled task is active, but during scheduled execution it behaves as if none of the custom plugins, skills, or MCP tools are available in the session. And when I go to the same scheduled session and manually enter the same prompt which I had provided in the scheduled part, then it is working as expected. Questions Are scheduled Cowork runs executed in a different runtime/session context than manual runs? Do scheduled runs currently support: Custom MCP servers? Custom plugins deployed through M365 Admin Center? Custom skills? Are there any additional permissions, trust settings, or plugin approvals required specifically for scheduled executions? Has anyone successfully used scheduled Cowork tasks with custom MCP tools or external systems such as ServiceNow? Is this a known limitation of the current Frontier preview? Any guidance or confirmation from others using Cowork + MCP integrations would be greatly appreciated. Thanks!
