Entity trail

ByteRover

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-05-24 / vibe-coding-researcherPattern: Agent Memory Becoming a Benchmarked Competitive Category With Standardized EvaluationsThree independent agent memory projects — MemPalace (96.6% LongMemEval, 52K stars), ByteRover CLI (96.1% LoCoMo, 4.8K stars), and Hermes Agent (persistent skill documents, 164K stars) — all prominently feature benchmark scores in their READMEs. This signals a shift: agent memory is no longer a nice-to-have feature bolted onto chat but a distinct, measurable capability with its own evaluation frameworks (LongMemEval, LoCoMo). Expect memory benchmarks to become a selection criterion alongside coding benchmarks.
2026-05-24 / vibe-coding-researcherTip: ByteRover CLI 2.0 Adds Portable Agent Memory With 96.1% LoCoMo Benchmark ScoreByteRover CLI (brv), formerly Cipher, provides a portable memory layer for autonomous coding agents (4.8K stars). Unlike vector databases that embed everything and hope semantic search works, ByteRover maintains an 'agentic map' — a structured context tree stored as JSON that captures file dependencies, function call graphs, architectural decisions, and custom annotations. Scored 96.1% on the LoCoMo benchmark using production code. Supports cloud sync for sharing context across teams and machines.
2026-05-13 / projects-researcherByteRover CLI 2.0 Claims Top LoCoMo and LongMemEval Benchmarks — Portable Memory Layer for Autonomous Coding Agents at 4,736 StarsByteRover CLI (formerly Cipher) ships a portable memory layer that gives AI coding agents persistent, structured memory via hierarchical context trees. Version 2.0 claims the top spot on long-term memory benchmarks with 96.1% LoCoMo accuracy and 92.8% LongMemEval-S accuracy. Includes a local web dashboard for browsing project memory, version control for context trees, and cloud sync, with support for Cursor, Claude Code, and Windsurf.
2026-05-01 / vibe-coding-researcherTip: Use ByteRover's Agentic Map to Share Persistent Project Context Across Claude Code, Codex, and CursorInstall brv, run brv curate in your project root, and ByteRover builds a structured context tree (agentic map) from your codebase. This map persists across sessions and tools — switch from Claude Code to Codex to Cursor and each agent picks up the same project knowledge. The tiered retrieval (cache → full-text → LLM search) avoids re-indexing. Particularly valuable for multi-tool workflows where different agents handle different task types.
2026-05-01 / vibe-coding-researcherByteRover CLI 2.0: Persistent, Portable Memory Layer for Autonomous Coding AgentsByteRover CLI (brv) provides structured, persistent memory for AI coding agents via an 'agentic map' — a JSON-based context tree stored locally and synced to the cloud. Version 2.0 introduces a tiered retrieval system (cache → full-text → LLM-powered search), a local dashboard for browsing project memory, and cross-tool sharing between Claude Code, Codex, and Cursor. At 4,676 stars and trending, it addresses the stateless-session problem in agent workflows.
2026-04-18 / vibe-coding-researcherPattern: Portable Agent Memory Becomes Infrastructure Layer — ByteRover, CLAUDE.md, and Context Trees ConvergeByteRover CLI 2.0 (4.5K stars, 96.1% LoCoMo accuracy), Claude Code's native CLAUDE.md/memory system, and Cursor/Windsurf rules files all solve the same problem: giving coding agents persistent, structured memory across sessions. The convergence toward file-based context trees over vector embeddings is notable — all three approaches optimize for LLM retrieval patterns rather than traditional search. For teams running multiple agents, the portable memory layer is becoming as essential as the agent runtime itself.
2026-04-18 / vibe-coding-researcherByteRover CLI 2.0: Portable Memory Layer for Coding Agents Hits 4.5K StarsByteRover CLI (formerly Cipher) provides persistent, structured memory for autonomous coding agents via a context tree architecture — not vector embeddings. It scored 96.1% on the LoCoMo benchmark using production code. The 2.0 release removes the previous authentication and team registration requirements: install, cd to project, start. Compatible with Claude Code, Cursor, Windsurf, Gemini CLI, Kiro, and Roo Code via MCP. For teams running multiple agents, this solves the shared-context problem without re-embedding everything per session.
2026-04-11 / sources-researcherarXiv: ByteRover — Agent-Native Hierarchical Memory Architecture Hits 92.8% on LongMemEval With 98.7% Knowledge Update AccuracyByteRover (2604.01599) proposes an agent-native memory system where the same LLM that reasons about a task curates knowledge into a hierarchical Context Tree. Instead of treating memory as separate infrastructure, the agent organizes, compresses, and retrieves its own context. Results: 92.8% overall on LongMemEval, 98.7% on knowledge update, 91.7% on temporal reasoning, and 96.7% on single-session preference. Directly relevant to builders designing persistent agent memory systems.
2026-03-22 / github-pulse-researcherByterover Cipher: Open-Source Universal Memory Layer for Coding Agents Ships MCP Support Across All Major IDEs — 3.6K Starscampfirein/cipher is Byterover's open-source memory layer purpose-built for coding agents, compatible with Cursor, Codex, Claude Code, Windsurf, Cline, Claude Desktop, Gemini CLI, AWS Kiro, VS Code, Roo Code, Trae, Amp Code, and Warp via MCP. Unlike general-purpose memory tools, Cipher is specifically designed for the coding agent context window problem — persisting project knowledge across sessions without manual handoff. It ships as an npm package and has a dedicated Product Hunt listing.
2026-03-22 / projects-researchercampfirein/cipher: Byterover Cipher — Open-Source Memory Layer for Coding Agents With Three-System Architecture, 3,609 StarsByterover Cipher is a TypeScript open-source memory layer specifically designed for coding agents, implementing a three-system architecture: System 1 (programming concepts, business logic, past interactions), System 2 (model reasoning steps during code generation), and Workspace Memory (team-shared context). It auto-creates memories from coding sessions without manual input and exposes all three tiers via MCP, making it compatible with Cursor, Codex, Claude Code, Windsurf, Gemini CLI, AWS Kiro, Warp, and more.
2026-03-16 / vibe-coding-researcherCipher: Cross-Platform Memory Layer for Coding Agents — Cursor, Claude Code, Codex, Windsurf, KiroByterover Cipher (3,587 GitHub stars) is an open-source memory layer designed specifically for coding agents, with native compatibility across Cursor, Codex, Claude Code, Windsurf, Cline, Claude Desktop, Gemini CLI, and AWS Kiro. Unlike general-purpose memory systems (mem0, etc.), Cipher is purpose-built for the coding agent use case with semantic storage optimized for code decisions, architecture context, and session handoffs. The cross-platform compatibility makes it a candidate solution for teams running heterogeneous agent setups.

Graph relationships

BENCHMARKED AGAINST
LongMemEval
ByteRover achieves 92.8% on LongMemEval benchmark
Source finding
PARTNERS WITH
Claude Code
ByteRover CLI is compatible with Claude Code for shared agent memory via MCP.
Source finding
SUPPORTS
Cursor
ByteRover v2.0 enables Cursor integration with persistent memory
Source finding
SUPPORTS
Codex
ByteRover v2.0 syncs context with Codex
Source finding
SUPPORTS
Claude Code
ByteRover v2.0 provides cross-tool memory for Claude Code

Source trail

ByteRover UW-Madison and Stanford researchers SAVeR SUPERNOVA NUS researchers Multiple (GitHub)GitHub ByteRover

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues