TIZ36/evo

README 摘要

evo Evolving memory for agents — recall before the model thinks, reflect after it answers. evo is a provider-neutral memory service with a Cordis plugin and a native DeepSeek Harness adapter. It recalls durable memory into the system prompt before a model step, and reflects completed turns into structured memory afterwards — so an agent keeps what it learned across sessions without you writing it down. Renamed: this project was published as evo-memory until 0.2.x . The old data directory, the old EVO MEMORY environment variables, and the old /evo-memory/ HTTP prefix all still work — see Migrating from evo-memory . Contents - Features · Requirements · Install · Quick start - How it works · Workspace import · Settings panel - Claude Code hook · Codex plugin · HTTP API - Configuration · Service API · Scopes - Privacy · Limitations · Development · License Features - Recall + reflect loop. Memory is assembled into the system prompt on every turn, and each successful turn is distilled back into structured memory items. - Structured, not a scratchpad. Every item has a scope, a kind ( fact , constraint , procedure , skill , …), tags, and a source pointing at the session and turn it came fr…

分类