Memoir FAQs

Question 1

What kind of information can Memoir log for an AI agent?

Accepted Answer

Memoir offers a comprehensive suite of tools to `log_attempt`, `log_blocker`, `resolve_blocker`, and `log_decision`. It also provides `get_handoff`, `get_history`, and `get_blockers` to retrieve past session information.

Question 2

Is Memoir difficult to set up or configure?

Accepted Answer

No, Memoir is designed for ease of use. It features automatic project detection via git root and creates sessions implicitly on the first log, requiring no explicit setup. All limits are also configurable via environment variables.

Question 3

Where does Memoir store its session data?

Accepted Answer

Memoir stores all session data locally in a single SQLite file located at `~/.memoir/memoir.db`. It requires no API keys, cloud services, or external connections, ensuring your data remains private and self-contained.

Question 4

How does Memoir manage old session data?

Accepted Answer

Memoir uses rolling retention, automatically pruning old sessions to keep a maximum of 20 sessions per project by default. This limit, along with others like content and entry limits, is fully configurable via environment variables.

Question 5

What is Memoir and how does it help AI agents?

Accepted Answer

Memoir provides persistent, structured session journals for AI agents. It logs attempts, blockers, and decisions, enabling agents to remember past experiences, prevent repeating mistakes, and facilitate continuous learning across sessions.

Memoir

Memoir

Key Features

Use Cases

Key Features

Use Cases