CodeBurn
Understand coding-agent spending through local session, model, and project breakdowns.
Keep searchable local memory across AI coding sessions and tools

agentmemory is an open-source memory service for AI coding tools. It captures prompts and tool activity from supported agents, stores observations locally, and retrieves relevant context when a later session needs it. A shared local server lets different coding assistants use the same history rather than starting with separate project knowledge.
The system indexes observations, consolidates related information, and selects a compact set of relevant memories for the next task. Hooks can restore context around session starts and compaction, while search supports questions about earlier changes or decisions. Its default configuration binds to the local machine with authentication enabled and does not require a cloud account; optional provider connections change that local-only arrangement.
Tried it? Share a useful thought or ask a question.
Good conversations start with a hello.
Sign in to comment