Zaxy — event-sourced agent memory with cited checkout and reversible forgetting (stdio/SSE via PyPI)

What is Zaxy?

Zaxy is a production memory server for agent teams that need auditable, event-sourced context management. It captures parent missions, worker sessions, tool observations, cited findings, conflict review, approval packets, and merge-back into a durable, queryable history. Under the hood it uses Eventloom append‑only JSONL as the source of truth and an embedded LadybugDB graph projection for local reasoning.

How to use Zaxy?

Install the PyPI package zaxy-memory with pipx install zaxy-memory, then run zaxy init to set up a local embedded graph, repo‑local profile, and deterministic capture config. Verify with zaxy memory log and zaxy doctor. Start the MCP server over stdio or SSE with zaxy serve. Run the single‑agent example with python examples/single_agent_memory.py. For Claude Code, use zaxy init --preset local-claude; for Hermes Agent, zaxy ide-config hermes --install.

Key features of Zaxy

• Immutable audit trail via Eventloom append‑only JSONL with SHA‑256 hash chains.
• Bi‑temporal graph with validity windows (valid_from, valid_to).
• Hybrid extraction: rule‑based for typed events, LLM fallback.
• Hybrid retrieval: exact + keyword + vector + graph traversal with configurable fusion weights.
• Session sharding: one Eventloom log per agent/session, shared graph.
• MCP‑native integration over stdio or SSE.
• Optional Pathlight traces, breakpoints, and diff support.
• Hardened local defaults: bounded inputs, safe session IDs, embedded graph without sidecar.

Use cases of Zaxy

• Auditable memory for agent teams: every accepted fact points back to Eventloom history.
• Agent‑team coordination: parent and worker sessions stay isolated until findings are reviewed and merged.
• Local‑first runtime with optional Neo4j sidecar for production scale.
• MCP‑native drop‑in memory for Codex, Claude Code, Cursor, VS Code, Hermes Agent, LangGraph, CrewAI, and AutoGen.

FAQ from Zaxy

What storage does Zaxy use?

Zaxy uses Eventloom append‑only JSONL files (one per session) as the immutable log, and an embedded LadybugDB graph projection for local real‑time reasoning. Optionally, it can also use a Neo4j sidecar via zaxy-memory[neo4j].

Does Zaxy require Neo4j?

No. The default installation uses the embedded LadybugDB graph. The Neo4j backend is optional and installed with zaxy-memory[neo4j].

What happened to the benchmark claims?

The prior LongMemEval numbers have been withdrawn because they were produced in oracle mode (pre‑selected gold sessions) and are not a measure of full‑haystack retrieval. Zaxy does not currently publish a LongMemEval score.

What transports does Zaxy support?

Zaxy supports both stdio and SSE (Server‑Sent Events) transports for MCP communication.

What is the PyPI package name?

The PyPI distribution is zaxy-memory (the name zaxy is taken). The import package and console command remain zaxy.