AI Chat

PeopleClaw provides a powerful conversational AI experience across both Telegram and the web dashboard. Here's how it works.

How Conversations Work

When you send a message, your AI assistant processes it within your dedicated sandbox environment. The assistant maintains conversation memory — it remembers what you've discussed within the current session, so you can have natural, multi-turn conversations.

Your assistant can also execute code, browse the web, generate files, and use installed skills — all within a single conversation.

Context Memory

PeopleClaw maintains context across messages in a conversation. The assistant remembers previous messages, so you can:

• Ask follow-up questions without repeating context
• Build on previous responses
• Reference earlier parts of the conversation

PeopleClaw also stores long-term memory — important facts, preferences, and context persist across conversations. Your assistant learns your preferences over time.

Resetting Conversations

To start a fresh conversation and clear the current context, use the /reset command. This clears the conversation history but preserves your long-term memory and settings.

Streaming Responses

On the web dashboard, responses stream in real-time — you see the text appear as it's being generated, similar to ChatGPT.

On Telegram, the assistant sends the complete response once it's ready. For long responses, the message may be split into multiple parts.

Web Chat vs Telegram

Tips

• Be specific in your requests for better results
• Use /reset when switching to a completely different topic
• Upload files directly for analysis — the assistant can read images, documents, and more
• Try different models for different tasks — coding models excel at code, reasoning models at complex analysis