Skip to content

first-conversation.md

Latest commit

 

History

History
72 lines (49 loc) · 1.93 KB

first-conversation.md

File metadata and controls

72 lines (49 loc) · 1.93 KB

First Conversation

Chat with your configured agent using the interactive CLI.

Quick Start

hexis chat

This opens an interactive conversation loop with memory enrichment and tool access.

How It Works

The chat loop automatically:

  1. Enriches your prompt with relevant memories from the agent's brain (RAG-style)
  2. Gives the agent tools via function calling -- memory operations, web search, file access, and more
  3. Forms new memories from the conversation (the agent remembers what you discuss)

Chat Options

# Default: memory tools + extended tools (web, filesystem, shell)
hexis chat

# Specify a different LLM endpoint
hexis chat --endpoint http://localhost:11434/v1 --model llama3.2

# Memory tools only (no web/filesystem/shell)
hexis chat --no-extended-tools

# Quiet mode (less verbose output)
hexis chat -q

What to Try

  • Ask about itself -- "What do you know about yourself?" (tests identity retrieval)
  • Tell it something -- "I prefer dark mode" (tests memory formation)
  • Ask a follow-up -- "What do I prefer?" (tests memory recall)
  • Give it a goal -- "I want you to help me learn Python" (tests goal creation)

Understanding the Output

During chat, you may see:

  • Memory recall indicators -- shows which memories were retrieved for context
  • Tool calls -- shows when the agent uses tools (recall, remember, web_search, etc.)
  • Memory formation -- indicates new memories being created from the conversation

Verify Memories Were Created

After chatting, check that the agent remembered:

hexis recall "what we discussed"    # search memories
hexis status                         # see memory counts

Next Steps