Manus

@manus list all tasks in the Q1 Product Launch project and show me which ones are still in progress

Based on the Q1 Product Launch project (ID: proj_abc123), here are the in-progress tasks: 1. **Feature specification review** - Started 2 days ago, 60% complete 2. **API integration testing** - Started yesterday, 25% complete 3. **Marketing copy draft** - Started 4 hours ago, 15% complete All three tasks are actively processing. The feature spec review should complete within the next few hours based on current progress rate. Would you like me to check detailed status or attachments for any of these?

This example shows Manus's project-based task organization. The MCP retrieves task metadata including progress percentages and timing. Note that you'll need valid project_id values from your Manus account—projects must be created first via the Create New Project tool before tasks can be queried under them.

@manus create a new task to analyze our competitor pricing from the uploaded spreadsheet and generate a summary report

Task created successfully (ID: task_xyz789). I've initiated an AI task that will: - Process the pricing spreadsheet you uploaded (file_def456) - Extract competitor pricing data across product tiers - Generate a comparative analysis summary The task is now running. Based on typical processing times for spreadsheet analysis, expect results in 3-5 minutes. I've set up the task to notify via webhook when complete. You can continue this conversation by referencing task_xyz789 for follow-up questions or refinements.

This demonstrates Manus's core workflow: creating AI tasks with file attachments. The Create New Task tool supports multi-turn conversations via taskId, letting you refine outputs iteratively. Important: file uploads require a two-step process (Create File for presigned URL, then PUT to S3) before attaching to tasks.

@manus summarize the outputs from all completed tasks in the last 48 hours and identify any common themes or blockers mentioned

I've reviewed 12 completed tasks from the past 48 hours. Here are the key themes: **Common blockers (mentioned in 7 tasks):** - API rate limiting on external data sources - Missing stakeholder approval on design specs **Recurring topics:** - Customer feedback analysis (4 tasks) - Technical documentation updates (3 tasks) - Competitive research (2 tasks) Three tasks flagged the API rate limit issue as causing delays. The stakeholder approval blocker appears in both the design and marketing task streams. Would you like me to create a consolidated report task that digs deeper into the rate limit patterns?

This synthesis example pairs Manus's task retrieval with AI reasoning to surface patterns across multiple task outputs. The MCP provides raw task data; the AI interprets it. Useful for team retrospectives or identifying systemic issues. Keep in mind that task history depth depends on your Manus plan's retention limits.

When Manus wins for client deliverable review loops

A 6-person creative agency ships 20+ client deliverables a week—landing pages, email templates, social assets. Each needs a QA pass before handoff. Manus is the right call here because you can spin up a project per client with standing instructions (brand guidelines, accessibility checks, file naming rules), then create tasks for each deliverable. The file upload flow (Create File + presigned S3 URL) handles PDFs and images without choking your workspace. The webhook tool means your PM gets a Slack ping when QA finishes, no manual polling. The threshold: if your agency runs on Figma comments or needs live annotation, Manus won't replace that—it's async task orchestration, not real-time collaboration. But for structured, repeatable QA at volume, it keeps the work moving without a dedicated QA Slack channel.

When Manus handles one-off research tasks for support teams

A 3-person support team at a B2B SaaS company fields 40 tickets a day. Half require digging through old Zendesk threads, Notion docs, or Slack archives to answer. Manus works well for the one-off research tasks that don't fit a canned macro. You create a task with the customer question, attach relevant files (past tickets, internal docs), and let the AI synthesize an answer. The multi-turn conversation support (taskId continuation) means you can refine the output if the first pass misses context. The trade-off: if your support flow needs sub-5-minute response times, Manus adds latency—you're waiting on task completion, not instant chat. It shines when the alternative is a support engineer spending 20 minutes hunting down an answer. For teams drowning in context-heavy tickets, Manus turns research into a delegated task.

When Manus scales content ops without hiring a coordinator

A 4-person content team publishes one long-form blog post a week, then needs to repurpose it into a Twitter thread, LinkedIn post, email snippet, and meta description. Manus is the right tool because you can set up a project with repurposing instructions, then create a task per output format. The file upload tool handles the source doc (Markdown, Google Doc export, PDF), and the task system queues the work so you're not babysitting five chat threads. The webhook integration means your editor gets notified when all variants are ready for review. The boundary: if your repurposing needs heavy brand voice tuning or you're publishing 10+ posts a week, you'll outgrow the task-per-output model—you'd want a dedicated content ops platform. But for teams doing 4-8 repurposing jobs a week, Manus removes the coordination tax without adding headcount.