# Make something useful with Workspace.

Reviewed: 2026-09-12

Canonical: https://clankercloud.ai/learn

Step-by-step guides to Projects, source verification, CSV cleanup, reusable Agents, private Apps, cloud browsers, and recurring work. Try a worked example and check your result.

## Projects and files

### [Your rough notes. A report worth sending.](https://clankercloud.ai/learn/turn-notes-into-a-report)

Create a client report from meeting notes and Project files. Review progress, decisions, blockers, and next steps before downloading or sharing.

### [Use your Google files. Keep the context.](https://clankercloud.ai/learn/google-drive-docs-sheets)

Connect selected Google files to Workspace, edit Docs and Sheets from a Session, and save Project copies. Understand experimental access and copy limits.

### [Give your Project a source of truth.](https://clankercloud.ai/learn/organize-project-files)

Organize AI Project files with clear versions, source notes, and review rules. Prevent stale inputs from becoming confident answers in later Sessions.

### [Find what changed and why it matters.](https://clankercloud.ai/learn/compare-document-versions)

Compare two document versions with source locations, material-change categories, and unresolved questions. Review a fictional project-scope example.

### [Keep the result and the context needed to trust it.](https://clankercloud.ai/learn/save-and-reuse-results)

Choose Project files, Bookmarks, Wiki articles, or reusable Agents for finished AI work. Preserve sources, limits, and refresh rules for the next task.

## Agents and access

### [One familiar job. Your own specialist.](https://clankercloud.ai/learn/build-a-custom-agent)

Create a reusable AI Agent with instructions, reference files, Skills, and selected Tools. Test it on a real example before reusing or sharing.

### [Write instructions your Agent can prove it followed.](https://clankercloud.ai/learn/write-reusable-agent-instructions)

Write and test AI Agent instructions using input rules, output criteria, exceptions, and known examples. Separate reusable methods from changing Project facts.

### [Give an Agent the right knowledge, method, and operations.](https://clankercloud.ai/learn/agent-knowledge-skills-tools)

Understand how Agent knowledge, Skills, and Tools work together. Configure a proposal-review specialist and diagnose missing context or unnecessary access.

### [Choose approvals around the actions you intend.](https://clankercloud.ai/learn/choose-approval-mode)

Choose Workspace approval settings by action scope. Review Agent capabilities, test changes, and separate Session approvals from unattended worker access.

## Research and data

### [Compare your sources. Make a clearer decision.](https://clankercloud.ai/learn/research-with-project-files)

Compare proposals and research files in an AI Project. Build an evidence table, flag missing information, and save a recommendation you can review.

### [Check the claim. Open the source.](https://clankercloud.ai/learn/verify-ai-research)

Verify AI research with a claim register, source checks, conflicting evidence, and an explicit decision date. Review a fictional vendor-policy example.

### [Clean a CSV without losing what it means.](https://clankercloud.ai/learn/clean-csv-data)

Clean CSV data with explicit rules for blanks, dates, identifiers, and duplicates. Reconcile rows and totals before using the result in an App or report.

## Apps and sharing

### [A spreadsheet in. A useful App out.](https://clankercloud.ai/learn/build-an-app-from-a-spreadsheet)

Turn a spreadsheet workflow into an AI-built tracker or dashboard. Test a private App preview and review the data before sharing or publishing.

### [Test the workflow behind the preview.](https://clankercloud.ai/learn/test-a-private-app)

Test a private AI-built App with expected results, empty states, keyboard checks, and reload behavior. Work through a fictional three-record tracker.

### [Share a reviewed App with the intended people.](https://clankercloud.ai/learn/share-apps-privately)

Prepare an App for private sharing, inspect its data and access, and test the recipient experience. Understand how private sharing differs from publishing.

## Recurring work

### [Test it once. Give it a rhythm.](https://clankercloud.ai/learn/loops-and-schedules)

Set up recurring AI work with a Loop or Schedule. Understand timing, unattended computer jobs, history, cloud execution, and ongoing credit costs.

### [Start small. Know what keeps running.](https://clankercloud.ai/learn/workspace-credits-and-costs)

Understand welcome credits, optional plans, AI usage, and recurring computer costs. Estimate a Workspace task without assuming unlimited or fixed-price work.

### [Check what the worker did, then decide what runs next.](https://clankercloud.ai/learn/review-recurring-worker-results)

Review recurring AI worker history, saved outputs, source failures, and uncertain actions. Decide when to pause, repair, or run a task again.

## Computer work

### [Put website work in the right browser.](https://clankercloud.ai/learn/use-cloud-browsers)

Scope a cloud-browser task, verify the active account and website, review extracted facts, and stop the environment when the work is finished.

[Open Workspace](https://workspace.clankercloud.ai)

[Current pricing and usage](https://clankercloud.ai/pricing)