my AI agent forgets everything. so i'm spending october fixing that

Description and use

The author tackles AI agents’ single‑session memory by integrating the open‑source cognee memory layer and building new data‑source connectors, enabling persistent knowledge across sessions.

Problem

AI agents lose context after each session, requiring re‑explanation of prior decisions and data.

Outcome

By building a new connector for a real data source, the author demonstrates that the agent can remember and search across sessions, reducing the need for repetitive prompts.

Workflow

• Select a data‑source connector that can be tested on a real account
• Claim the connector issue and wait for assignment
• Implement the connector following the existing Notion connector pattern
• Enable incremental sync to pull only changed data
• Handle deletions so the agent forgets removed items
• Write tests, a README, and a runnable example
• Publish a blog post per PR describing real‑world usage

Evidence

most of the agents i build for clients are smart for exactly one conversation.

Source

cognee is an open source memory layer for AI agents. you feed it documents, chats, code and data, and it turns them into a knowledge graph your agent can actually remember and search across sessions.

Source

an agent only remembers what it can connect to

Source