memory-configure

name	memory-configure
description	Wizard-style configuration guidance for a single-agent agent-memory setup. Shows defaults inline, provides a full sample config, and includes a dry-run config check step. Always confirm before creating or editing files.
license	MIT
metadata	{"version":"1.0.0","author":"SpillwaveSolutions"}

Memory Configure Skill

Guide users through configuring agent-memory for a single-agent setup. This skill does not apply multi-agent configuration. Always ask for confirmation before creating or editing configuration files. Provide verification commands only; do not run them.

When to Use

• User needs a first-time config file
• User wants to reset to defaults
• User wants to update the summarizer provider/model

When Not to Use

• Installation (use memory-install)
• Verification checks (use memory-verify)
• Troubleshooting failures (use memory-troubleshoot)

Wizard Principles

• Ask one question at a time
• Confirm before any file creation or edit
• Provide verification commands only
• Keep scope to single-agent defaults

Defaults (single-agent)

• db_path: ~/.local/share/agent-memory/db
• grpc_port: 50051
• grpc_host: 0.0.0.0
• log_level: info
• summarizer.provider: openai
• summarizer.model: gpt-4o-mini

Wizard Flow

Step 1: Check for existing config

Do you already have a config file at:
  ~/.config/agent-memory/config.toml

1. Yes
2. No

Enter selection [1-2]:

If yes, ask if they want to keep it, overwrite, or edit.

Step 2: Choose summarizer provider

Which summarizer provider should be used?

1. OpenAI (gpt-4o-mini)
2. Anthropic (claude-3-5-haiku-latest)
3. Local (Ollama or compatible)
4. None (disable summarization)

Default: 1

Enter selection [1-4]:

If provider needs a key, remind user to set the environment variable:

OpenAI:   export OPENAI_API_KEY="sk-..."
Anthropic: export ANTHROPIC_API_KEY="sk-ant-..."

Step 3: Show the full sample config

# ~/.config/agent-memory/config.toml

db_path = "~/.local/share/agent-memory/db"
grpc_port = 50051
grpc_host = "0.0.0.0"
log_level = "info"
multi_agent_mode = "separate"

[summarizer]
provider = "openai"
model = "gpt-4o-mini"

Step 4: Confirm file creation or edit

I can create or update:
  ~/.config/agent-memory/config.toml

Proceed? (yes/no)

If yes, ask whether to create directories as needed, then proceed.

Step 5: Dry-run config check (optional)

Provide command only:

Optional verify command:
  memory-daemon config check

Step 6: Next steps

• Use memory-verify for health checks
• Start the daemon with memory-daemon start
• Configure agent hooks using the agent-specific setup guide

Confirmation Language

Any time a file change is requested, ask explicitly:

This will edit ~/.config/agent-memory/config.toml.
Proceed? (yes/no)

If the user says no, provide the commands and stop.