The AI coding agent that never forgets

OpenZero persists knowledge across sessions in a vector database and compresses old context without losing critical decisions. Self-hostable. Open source. MIT licensed.

$
Get started Releases

Built for precision

6 Typed memory schemas
<200ms Semantic retrieval
4096 Embedding dimensions
3-tier Context compression

Two engines, one goal

Most AI assistants silently truncate your history. OpenZero takes a fundamentally different approach with two systems that work together.

Structured memory

Extracts durable facts after every response. Six typed schemas — Workflow, Bug Fix, Architecture, Preference, Config, Fact — stored as 4096-dim vectors in Qdrant.

Hierarchical compression

A three-tier engine keeps recent messages at full fidelity while progressively summarizing older history. Achieves 60-80% compression without losing critical context.

Hook-driven lifecycle

Memory retrieval runs before each user message. Extraction and compression run after each assistant response. Neither blocks the live chat path.

Semantic retrieval

Before every message, cosine-similarity search finds relevant memories from your entire project history and injects them as structured context.

Per-user isolation

All memories are partitioned by hashed user ID. Every Qdrant query is filtered by this ID. No cross-user access. Metadata is scrubbed of PII during extraction.

Resilient pipeline

Three-level fallback on extraction and retrieval. Exponential-backoff retries. Neither subsystem can crash or block the primary chat loop.

Six schemas, zero guesswork

Every extracted memory is classified into a strict type so retrieval understands the nature of the knowledge, not just its text similarity.

Workflow

Repeatable commands and operational processes. Tracks trigger conditions and dependencies.

Bug Fix

Solved problems and post-mortems. Symptom, root cause, solution, and prevention guidance.

Architecture

Design decisions with rationale, alternatives considered, and known tradeoffs.

Preference

Coding styles, naming conventions, and tool choices. Concrete examples included.

Config

Environment settings and infrastructure variables. Name, value, location, and purpose.

Fact

General project knowledge that doesn't fit other types. Free-form with semantic keywords.

Stop re-explaining things to your AI

Install OpenZero in under a minute. Requires Bun, Docker for Qdrant, and an LLM provider key.