aiqbench analysis
This paper introduces `memorywire`, an open-source vendor-neutral wire format designed to standardize agent memory operations across disparate frameworks like mem0 and Letta. By defining a common protocol for semantic, episodic, procedural, and emotional memories—including optional human-in-the-loop governance—it directly addresses the current fragmentation that forces bespoke integrations and rebuilds during migration. The work focuses on practical interoperability rather than novel algorithms, offering immediate utility for developers seeking to compose memory systems without being locked into specific storage layouts or SDKs.
memorywire: A Vendor-Neutral Wire Format for Agent Memory Operations
arXiv:2606.01138v3 Announce Type: replace-cross Abstract: Agent-memory frameworks -- mem0, Letta/MemGPT, Cognee, Zep/Graphiti, MemoryOS, MemTensor -- each ship their own SDK, storage layout, and operational vocabulary. There is no shared wire format: every integration is bespoke, every migration rebuilds memory from scratch, and no framework ships a governance surface that lets a human review writes before they enter long-term storage. We present memorywire, a JSON-Schema 2020-12 wire format for five memory operations (remember, recall, forget, merge, expire) over four memory types (semantic, episodic, procedural, emotional), with a MemoryStore interface, a fan-out router, and an optional HITL governance channel. We describe an open-source reference implementation with five backend adapters (sqlite-vec, mem0, Letta, Cognee, pgvector); a microbenchmark on a 100-fact / 50-query labelled corpus (42 with non-empty gold ids + 8 no-match probes) achieving recall@5 = 1.000 on the 42 gold-id queries with ingest p50 = 37.8 ms and recall p50 = 40.6 ms; an adversarial-fusion experiment showing Reciprocal Rank Fusion holds recall@5 = 1.000 across a 1-of-N rank-0 injection sweep (K in {0, 5, ..., 50}) where max fusion collapses to 0.500 with 80% leak at K >= 5; and a 16-scenario cross-adapter conformance suite passing 68 of 80 cells with zero failures. The contribution is not a new algorithm; it is a packaging of established components (RRF, FSMs, STM/LTM consolidation, diff-and-approve workflows) into a venue-neutral protocol with an empirically validated reference, positioned to compose with the Model Context Protocol rather than compete with it. We further show that memorywire's provenance field is the strongest lever for recovering a poisoned store, evaluated with an external benchmark (PurgeBench).
Abstract:Agent-memory frameworks -- mem0, Letta/MemGPT, Cognee, Zep/Graphiti, MemoryOS, MemTensor -- each ship their own SDK, storage layout, and operational vocabulary. There is no shared wire format: every integration is bespoke, every migration rebuilds memory from scratch, and no framework ships a governance surface that lets a human review writes before they enter long-term storage. We present memorywire, a JSON-Schema 2020-12 wire format for five memory operations (remember, recall, forget, merge, expire) over four memory types (semantic, episodic, procedural, emotional), with a MemoryStore interface, a fan-out router, and an optional HITL governance channel. We describe an open-source reference implementation with five backend adapters (sqlite-vec, mem0, Letta, Cognee, pgvector); a microbenchmark on a 100-fact / 50-query labelled corpus (42 with non-empty gold ids + 8 no-match probes) achieving recall@5 = 1.000 on the 42 gold-id queries with ingest p50 = 37.8 ms and recall p50 = 40.6 ms; an adversarial-fusion experiment showing Reciprocal Rank Fusion holds recall@5 = 1.000 across a 1-of-N rank-0 injection sweep (K in {0, 5, ..., 50}) where max fusion collapses to 0.500 with 80% leak at K >= 5; and a 16-scenario cross-adapter conformance suite passing 68 of 80 cells with zero failures. The contribution is not a new algorithm; it is a packaging of established components (RRF, FSMs, STM/LTM consolidation, diff-and-approve workflows) into a venue-neutral protocol with an empirically validated reference, positioned to compose with the Model Context Protocol rather than compete with it. We further show that memorywire's provenance field is the strongest lever for recovering a poisoned store, evaluated with an external benchmark (PurgeBench).
18 pages, 1 figure, 6 tables. v3 adds a memory-poison recovery evaluation (PurgeBench) and a recover capability; provenance-based forget is the strongest recovery lever, the entangled class remains open. Code: this http URL. Companion to arXiv:2604.18248 (Prompt Injection Detection)
Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC)
arXiv:2606.01138 [cs.CR]
(or arXiv:2606.01138v3 [cs.CR] for this version)
Full text under the publisher's open licence · 320 words
Related coverage
memorywire: A Vendor-Neutral Wire Format for Agent Memory Operations
arXiv · Jun 4, 2026
Presentation: Architecting AI Systems for the Messy Reality of Enterprises: Why Agentic Compute is the Missing Layer
InfoQ · Aug 3, 2026
Spur: Modern GPU Job Scheduling for HPC and AI Workloads
AMD · Jul 22, 2026
Article: The Agent Harness: What it is and Two Ways to Build One
InfoQ · Sep 25, 2026