Skip to content

Quickstart

This guide takes you from a running Snippbot instance to your first conversation with a custom agent.


Open http://localhost:5173 in your browser. You should see the Snippbot dashboard with the sidebar navigation. If you see the login page, sign in with the credentials you created during the setup wizard.

If you see the setup wizard instead, complete it first — it walks you through system checks, identity, account creation, and AI provider configuration.


Before creating an agent, you need at least one AI provider API key configured. If you already set this up during the setup wizard (Step 5: Model), skip to Step 3.

  1. Open Settings

    Click Settings in the sidebar (bottom of the navigation under Admin).

  2. Go to AI Providers

    Select the Providers section in the settings sidebar.

  3. Enter your API key

    Click any provider to expand it and enter your API key:

    ProviderKey prefixWhere to get it
    Anthropicsk-ant-console.anthropic.com
    OpenAIsk-platform.openai.com
    OpenRoutersk-or-openrouter.ai
    GeminiAI...aistudio.google.com
    DeepSeeksk-platform.deepseek.com
    Groqgsk_console.groq.com
    Mistralconsole.mistral.ai

    After entering a key, the provider shows a green “Connected” badge.


  1. Start the Awakening

    Click Create Snippbot (the button at the top of the sidebar). This opens the Quick Awakening flow — a guided experience for creating a new agent.

  2. Choose a name

    Give your agent a display name. This is how it appears in the chat interface and agent list.

  3. Pick an avatar

    Select or upload a profile image for your agent.

  4. Select archetypes

    Choose one or more personality archetypes that define your agent’s behavior:

    ArchetypeWhat it’s good at
    Personal CompanionEmotional support, casual conversation
    Professional AssistantOrganization, task management
    ResearcherDeep research, knowledge synthesis
    Teacher / MentorExplaining concepts, guided learning
    Creative PartnerBrainstorming, artistic collaboration
    Coach / MotivatorAccountability, motivation, pushing toward goals
    Technical ExpertCode review, architecture, debugging
    Entrepreneur PartnerBusiness strategy, idea validation, MVP building
    Health / WellnessHealth, habits, mindfulness support
    Accountability PartnerKeeps you on track, follows up, celebrates wins
    Thinking PartnerThink through complex problems, brainstorm solutions
    Something UniqueYou describe exactly who you need

    You can select multiple archetypes to blend their traits.

  5. Review your agent’s profile

    Based on your chosen archetypes, Snippbot generates a set of personality attributes (communication style, expertise areas, etc.). Review and optionally edit these before continuing.

  6. Select a voice

    Choose a text-to-speech voice for your agent. You can preview different voice options and pick the one that fits your agent’s personality.

  7. Answer the getting-to-know-you questions

    The Awakening asks up to 5 personalized questions based on your chosen archetypes. Your answers shape the agent’s personality and communication style.

  8. Choose an evolution preference

    Decide how your agent should grow over time:

    • Evolve freely — adapt based on your interactions
    • Stay focused — remain specialized in the chosen role
    • Check with me — ask before making personality changes
  9. Done!

    Your agent is created and ready to chat. You’ll be redirected to the chat interface.


  1. Select your agent

    If you have multiple agents, click the agent selector at the top of the chat to choose which one to talk to.

  2. Send a message

    Type anything in the message input at the bottom and press Enter (or click Send). Your agent responds using the AI provider you configured. Use Shift+Enter for a new line without sending.

    Try something like:

    What are the top 5 open-source AI agent frameworks? Compare them in a table.
  3. Explore the tools activity

    The right panel shows a Tools Activity timeline — you can see what tools the agent uses (web search, file write, etc.) in real time as it works.


Behind the scenes, Snippbot:

  1. Created an agent with a workspace directory containing personality files (PERSONA.md, PROFILE.md, etc.)
  2. Routed your message to the selected AI provider using the agent’s configured model
  3. The agent processed your request, optionally using tools (search, file write, etc.)
  4. Streamed the response back to you token-by-token via SSE
  5. Stored the conversation in the agent’s episodic memory for future context