# What can you do in Clanker Cloud Workspace? Reviewed: 2026-10-25 Canonical: https://clankercloud.ai/learn/what-can-clanker-workspace-do Clanker Cloud Workspace is a browser-based AI workspace and agent factory. Startups, operators, VC teams, and small businesses can use Project files to research a decision, create and revise documents, analyze spreadsheets, prepare presentations, build private Apps, and keep tasks with their deliverables. Create reusable Agents with instructions, knowledge, Skills, and selected Tools, or connect a compatible external assistant through authorized Workspace MCP. Review the saved work and choose which actions and connections to enable. ## What you are working toward A useful starting workflow, the saved result you expect, and a clear path to a reusable Agent. ## Start with - One job with an intended reader or user - Approved source files and a definition of a good result - A reviewer, enabled tools, and a usage budget ## Choose the result you need Start with the deliverable rather than a long prompt. The table connects common business jobs to work you can inspect. Features depend on your account, enabled tools, permissions, and any required computer or service connection. | Job | Bring | Save and review | | --- | --- | --- | | Research a decision | Requirements, documents, and dated web sources | An evidence table with source links, uncertainty, and a decision brief. | | Write and revise a Word document | An approved brief or existing .docx | A saved .docx; check wording, tracked changes, comments, and layout. | | Analyze a spreadsheet | A dated CSV or workbook and metric definitions | A revised workbook or CSV; check formulas, totals, units, and missing rows. | | Prepare a presentation | An approved narrative, numbers, and an existing deck when revising | A saved slide deck; inspect every slide, chart label, and source. | | Make an internal work tool | A small approved dataset and a user task | A private App or HTML preview; test the inputs, calculations, and recipient access. | | Keep track of work | A task, Project, due date, and review owner | A Planner task with notes and exact saved deliverables. Assignment does not start execution. | | Reuse a specialist | A role, stable method, examples, and necessary operations | A custom Agent tested on a normal case, a missing input, and a correction. | | Repeat a checked task | Fresh-input rules, a schedule, and a spending limit | A configured Loop or Schedule with reviewed runs and saved results. | | Work from another assistant | A compatible client and authorized Project access | Saved Workspace outputs and a visible task you can inspect or take over. | ## Keep the sources and the next version in a Project Projects group files, instructions, Sessions, and saved results. Use a Project for a client, operating process, or piece of work. State which source is current and which statements are only proposals. Keep stable audience and evidence rules in Project instructions; keep this month’s facts in this month’s files. A Session is the working conversation. A saved deliverable is a separate object you should open and check. Add the next period’s sources or a dated correction, then ask for a revision. Uploading a file does not establish a live connection to the system it came from. Sources: - [Organize Project files](https://clankercloud.ai/learn/organize-project-files) - [Save and reuse reviewed results](https://clankercloud.ai/learn/save-and-reuse-results) ## Work with documents, spreadsheets, presentations, and PDFs Use Workspace to create or revise a Word document, work with spreadsheet data and formulas, and prepare or revise slide presentations. Open the actual output in Preview and check it against the source. For revisions, identify the exact saved file and preserve the original when you need a comparison. Document review can record comments and accept or reject supported changes. Saving a review decision and applying it to a new file are separate steps. Rendering, conversion, extraction, and complex document work may need project compute. Inspect coverage and warnings, especially for scanned PDFs, inferred tables, or a document larger than the prepared page range. Sources: - [Create and review Office deliverables](https://clankercloud.ai/learn/create-office-deliverables-with-ai) - [Compare document versions](https://clankercloud.ai/learn/compare-document-versions) ## Make Agents that keep a tested method An Agent holds a role and reusable behavior. Give it instructions, reference knowledge, Skills for a method, and selected Tools for operations. A reporting Agent can reuse the same review rules across client Projects while each Project supplies current facts. Agent creation is configuration, not foundation-model training. Test the saved result before reusing the role. Include missing facts and a legitimate correction; record what passed and what needed repair. A shared Agent copy does not give its recipient your connected accounts or credentials. Marketplace publication and recurring execution are separate decisions. Sources: - [Create and reuse your own Agents](https://clankercloud.ai/learn/ai-agent-factory) - [Build a custom Agent](https://clankercloud.ai/learn/build-a-custom-agent) - [Test the Agent before reuse](https://clankercloud.ai/templates/agent-acceptance-test) ## Connect the task to the finished work Planner keeps durable tasks across Sessions, with Project and Agent assignment, due dates, notes, and deliverables. Attach a saved file version or a useful handoff note so another person can find the evidence. Creating or assigning a task does not launch an Agent. A completed status does not prove that a document is accurate. For repeated work, configure a Loop or Schedule after checking the manual workflow. Define how it obtains fresh inputs, which actions it may take, how you review each result, and when it should stop. A saved Project or Agent alone does not schedule work. Sources: - [Planner tasks and deliverables](https://clankercloud.ai/learn/ai-planner-tasks-and-deliverables) - [Loops and Schedules](https://clankercloud.ai/learn/loops-and-schedules) ## Turn a useful result into a private App Use Apps when someone needs to filter a comparison, explore a small dataset, or work through a repeated decision. Start from an approved spreadsheet or explicit rules. Test a known input, an empty input, and a boundary case before relying on the result. Choose recipient access deliberately. A working preview is not proof that another account can open it. Review private sharing separately from public publication, and check the account’s sharing policy. Sources: - [Build an App from a spreadsheet](https://clankercloud.ai/learn/build-an-app-from-a-spreadsheet) - [Test a private App](https://clankercloud.ai/learn/test-a-private-app) - [Share Apps privately](https://clankercloud.ai/learn/share-apps-privately) ## Use native Agents or bring a compatible external assistant Work directly in Workspace Sessions with a selected Agent, or connect a compatible ChatGPT, Claude, Hermes, or OpenClaw client through the documented Workspace connection. Available features depend on the client, authorization, installed tools, and selected Project. Keep source files and saved results in Workspace so the next authorized assistant has a handoff. An external client connection does not automatically configure a recurring worker or grant every tool. Start with a read-only status check, then request one named saved output. Inspect the visible task and verify its files; use the takeover controls when you need to continue the work yourself. Sources: - [External assistant setup and verification](https://clankercloud.ai/workspace-agents) - [Native versus external Agents](https://clankercloud.ai/learn/native-agent-vs-external-agent) ## Add connected services and computer work when the task needs them Connected accounts and MCP expose the operations you authorize. The experimental Google integration can select accessible Drive files, work with Docs and Sheets, and save Project copies. An exported copy is a snapshot. Check the exact supported connection and operations before expecting live CRM, accounting, or other business-system access. Cloud browsers, hosted environments, and explicitly connected computers provide additional execution paths. A normal Session cannot see a computer merely because you are using its browser. Choose the required environment, inspect its access and availability, and stop unused compute. Eligible accounts can configure Clanker email addresses and allowed senders for email-triggered Sessions. This can start work automatically after setup; it is separate from connecting a personal mailbox. Image generation and other newly introduced capabilities may require account availability. Confirm current eligibility in the FAQ. Sources: - [Google connection](https://clankercloud.ai/learn/google-drive-docs-sheets) - [Choose where an Agent runs](https://clankercloud.ai/learn/choose-ai-agent-environment) - [Current Workspace capabilities, access, and costs](https://clankercloud.ai/faq) ## Choose a bounded first task Workspace is useful when you want sources, a reusable method, saved outputs, and a visible handoff together. For a single quick question, a chat may be sufficient. For a deterministic event pipeline, compare workflow automation. For infrastructure operations, evaluate Clanker DevOps separately. Start with approved or fictional files. Check the account’s permissions, data handling, current pricing, and usage before adding sensitive sources or automatic actions. Welcome credits and paid allowances are budgets, not a guaranteed number of finished reports. A useful trial ends with a result you can inspect. Sources: - [Compare with ChatGPT Projects](https://clankercloud.ai/compare/clanker-workspace-vs-chatgpt-projects) - [Compare with workflow automation](https://clankercloud.ai/compare/clanker-workspace-vs-n8n) - [Current pricing](https://clankercloud.ai/pricing) - [Data handling and security](https://clankercloud.ai/security) ## Try this prompt Copy into a Workspace Session. Opening Workspace does not send it automatically. ```text Help me choose one useful workflow in Clanker Workspace. My job is [job], my source files are [files], and the reader needs [result]. Ask about missing inputs, then propose a Project structure, one named saved deliverable, the minimum tools, and a review checklist. Explain whether a reusable Agent, Planner task, private App, or recurring run would help after the first result is checked. ``` ## Before using the result - The workflow has a named saved output and a person who can check it. - Every required operation is available and authorized. - Missing facts, costs, and any automatic actions are explicit. ## Common questions ### What can I do in Clanker Cloud Workspace? Keep Project files and instructions together, research decisions, create and revise Word documents, spreadsheets and presentations, build private Apps, and keep Planner tasks with saved deliverables. Configure reusable Agents, connect compatible external assistants, and set up recurring work when needed. Available operations depend on your account, enabled tools, permissions, and required connections or computer environment. [Read more in the FAQ](https://clankercloud.ai/faq#workspace-capabilities) ### Can I make my own AI Agents in Workspace? Yes. Configure a role with instructions, reference knowledge, Skills, and selected Tools. Test its saved work on known inputs, missing facts, and corrections before reusing it across Projects. This configures an Agent; it does not train a foundation model. Sharing, publication, connected accounts, and recurring execution are separate choices. [Read more in the FAQ](https://clankercloud.ai/faq#workspace-agent-creation) ### Can ChatGPT or Claude use my Workspace? Compatible clients can connect through authorized Workspace MCP to the operations you permit. Verify client support, sign in, select the intended Project, and start with a read-only check. Request a saved result and inspect it in Workspace. The connection does not automatically grant every tool or start recurring work. [Read more in the FAQ](https://clankercloud.ai/faq#workspace-external-assistants) ## Sources - [Workspace product and saved-output previews](https://clankercloud.ai/workspace) - [Current Workspace capabilities, access, and costs](https://clankercloud.ai/faq) - [Agent configuration, testing, and reuse](https://clankercloud.ai/learn/ai-agent-factory) - [Authorized external assistants](https://clankercloud.ai/workspace-agents) ## Related - [An AI workspace for startups and founders.](https://clankercloud.ai/learn/ai-workspace-for-startups) - [An AI workspace for operators who keep work moving.](https://clankercloud.ai/learn/ai-workspace-for-operators) - [An AI workspace for VC research and portfolio work.](https://clankercloud.ai/learn/ai-workspace-for-vcs) - [Your small business. With an AI team behind it.](https://clankercloud.ai/learn/ai-workspace-for-business) - [Turn Project files into Office deliverables you can review.](https://clankercloud.ai/learn/create-office-deliverables-with-ai) - [Keep the task, progress, and deliverable together in Planner.](https://clankercloud.ai/learn/ai-planner-tasks-and-deliverables) - [Try an operations review with a workbook, brief, and task handoff.](https://clankercloud.ai/templates/operations-review-pack) - [Build AI Agents for the work your business repeats.](https://clankercloud.ai/learn/ai-agent-factory) - [An AI agent workspace for ongoing business work.](https://clankercloud.ai/learn/ai-agent-workspace) [Open Workspace](https://workspace.clankercloud.ai) [Current pricing and usage](https://clankercloud.ai/pricing)