Quickstart

Start with one simple job. You do not need to learn the whole product before getting value.

One quick note first: if a Protect your account prompt appears after you sign in, that is Alfrada OS offering two-factor authentication — it is safe to accept (Account Security).

The Best First Move

Ask Alfrada OS whether it can help you achieve the objective.

That sounds simple, but it is often the highest-leverage start. Alfrada OS can plan the path, tell you what context is missing, choose the right capabilities, and sometimes even handle basic in-app setup steps for you.

Objective-first start

I want to achieve [objective]. Can you help me get there? Tell me the best path forward, what you need from me, and what we should do first.

This works well because it lets Alfrada OS guide the workflow instead of forcing a new user to design it alone.

The Easiest Way To Start

Say three things:

• what you want done
• what context you have
• what output you want back

Beginner prompt pattern

Help me with [task]. Use [my notes, uploaded file, or website] if relevant. Give me [the exact output I want]. Keep it [short, detailed, executive-ready, or draft-first].

This is easier for first-time users because it focuses on the job and output, not internal product concepts.

Let It Check For Earlier Context

If this is not a brand-new task, ask Alfrada OS to search your earlier work before it starts.

Find an older file

Search my conversation history and find `syz.docx`. I think this was about 5 days ago. Pull the relevant context into this session and tell me what it was about.

Resume a previous project

Search my previous conversations for the last time we worked on [project or client]. Summarize where we left off, what files matter, and what the best next step is.

Let It Use A Playbook If One Fits

Alfrada OS can work from playbooks for real jobs such as deep research, finance and market monitoring, Google Workspace work, presentations, Canva design, live dashboards, document and data analysis, coding and code audit, scheduled automation, and creative media. The full set lives in the Playbook Library.

You do not need to name the playbook yourself. Just tell Alfrada OS the objective and ask it to use the most relevant workflow if one already exists.

Use an existing workflow

I want to achieve [objective]. If there is already a playbook or repeatable workflow that fits this job, use it and tell me the best next step.

Copy One Of These

Research brief

Research [company or market] and give me a short brief with the main facts, risks, and opportunities. Include sources and keep it easy to skim.

File help

I uploaded this file. Tell me what it contains, what stands out, and what I should do next. Then give me a clean summary.

Presentation start

Help me turn this topic into a presentation. Give me a strong slide-by-slide outline with the key message for each slide.

Where Your Work Shows Up

As Alfrada OS works, a Work Panel opens beside the chat. It has up to eight tabs, and each one appears only once it has something to show:

• To-dos — the agent's live plan, ticked off as it works
• Swarm — the plan and workers when a job runs as a multi-agent Swarm
• Markdown — written drafts you can edit inline
• Uploads — the source files you have added
• Generated Files — everything produced this session
• Presentations — slide decks
• Dashboards — live interactive reports
• Live — real-time surfaces like a working browser or Video Studio

For the full tour of inputs and outputs, see Files and Outputs.

Good First Jobs

• ask for a market brief with sources
• upload a document and ask for a summary plus next steps
• upload a CSV and ask for a cleaning plan
• ask for a presentation outline
• ask Alfrada OS to resume an older project from your history
• ask it to set up a recurring monitoring task once the workflow is clear

What Happens Next

• Alfrada OS may ask one or two follow-up questions if something important is missing
• if you upload files, they stay attached to the session
• reports, decks, and files land in the Work Panel tabs as they are created
• if Agent Intuition is on (the toggle lives at Settings → Agent Behavior → Intuition), Alfrada OS can also draw on memory and playbooks to shape the work
• you can keep refining the work with short follow-up requests

Two useful follow-ups once a job is moving: ask Alfrada OS to check back on something and it can wake the same conversation up again after 5 minutes to 24 hours (Wake Me), while jobs worth repeating become Automations, managed at Settings → Automations.

One Simple Tip

Do not worry too much about naming the right tool. Just describe the result you want.

One setting worth knowing about: next to the model picker there is an Effort row — Auto / Fast / Medium / Beast — that controls how thoroughly Alfrada OS double-checks its own work, separate from the model you choose. Leave it on Auto to start (Models and Modes).

Trust, Then Verify

Alfrada OS is built to reduce many of the usual LLM problems by combining planning, tools, files, memory, and working context.

Part of the answer to "will it be wrong?" is built in. On bigger tasks Alfrada OS writes a task-specific rubric, grades its own draft against it, and revises — and the check is not hidden: it appears as a visible grading card in the conversation. On Medium Effort that is one review-and-revise pass; on Beast, Alfrada OS keeps revising until the rubric clears and also runs a vision check over generated charts and images. Trivial turns are never graded.

Still, if something is important, high-stakes, or easy to get wrong, check it. The best pattern is to trust Alfrada OS to move the work forward and verify the parts that matter most.

When You Are Ready For More

• read Writing Better Prompts to get better results
• read How Alfrada OS Works if you want the deeper product model
• save good prompts as playbooks once you find workflows you want to repeat