HomeProjectsArticlesPK BoardAbout
Explore Projects

Beads

Active
GitHub Go MIT

Description

A memory upgrade for coding agents. Provides persistent contextual memory for Claude Code, Codex, and other coding agents to improve long-task consistency.

Key Features

  • Dolt-powered versioned SQL database — cell-level merge, native branching and remote sync
  • Dependency-aware task graph — replaces Markdown TODOs with graph structure supporting blocks/related/parent-child relationships
  • Agent-optimized output — JSON output format, dependency tracking and auto-ready task detection
  • Hash-based conflict-free IDs — uses bd-a1b2 format IDs to prevent merge conflicts in multi-agent workflows
  • Semantic memory compaction — auto-summarizes old closed tasks to save context window space
  • Hierarchical task structure — supports Epic → Task → Sub-task hierarchy for complex project management

Use Cases

💡 Provide cross-session persistent task memory for Claude Code, Codex and other coding agents
💡 Track task dependencies with graph structures in multi-agent collaborative projects
💡 Use Dolt version control to review task change history and decision records
💡 Accumulate searchable knowledge insights for projects via bd remember
💡 Isolate experimental task branches for open source contributors to avoid polluting main repo

Strengths & Limitations

✅ Strengths

  • • Actively maintained, recent updates
  • • High community interest (26.4k stars)
  • • Permissive open-source license (MIT)

Tags

coding-agent memory context go persistent

Categories

💻 Coding Agent 🧩 Agent Memory

Quick Start

curl -fsSL https://raw.githubusercontent.com/gastownhall/beads/main/scripts/install.sh | bash
cd your-project
bd init
bd create "Implement auth" -p 0
Visit GitHub

Related Projects

Agentic Stack

2.2k · Python
Active A

A portable .agent workspace that stores memory, skills, and protocols so coding agents like Claude Code, Cursor, and Windsurf can share durable knowledge.

coding-agentmemoryskills +2
  • · Portable .agent folder sharing memory and skills across 12+ coding agent harnesses
  • · Local data layer unifying harness activity monitoring, cron runs, token/cost estimates, and KPI dashboards
  • · Data flywheel converting approved runs into trace records, context cards, eval cases, and training-ready JSONL
View on GitHub

AgentMemory

27.1k · TypeScript
Active A

A persistent memory system for AI coding agents, designed around real-world benchmarks to preserve context across sessions.

memorycoding-agentpersistent-memory +1
  • · Cross-session persistent memory — AI coding agents preserve context across sessions, no re-explanation needed
  • · MCP protocol support — 53 MCP tools compatible with Claude Code, Copilot, Cursor and other agents
  • · Auto hook system — 12 automatic hooks save and retrieve memories during agent operations
View full details → View on GitHub

ByteRover CLI

4.9k · TypeScript
Normal B

ByteRover CLI provides persistent structured memory for autonomous coding agents. It features context tree management, git-like version control, and cloud sync, compatible with Cursor, Claude Code, Windsurf, and 22+ coding agents via MCP integration.

agent-memorycontext-managementcoding-agent +3
  • · Persistent structured memory with context tree and knowledge storage for coding agents
  • · Git-like version control for context trees (branch, commit, merge, push/pull)
  • · Supports 20 LLM providers including Anthropic, OpenAI, Google, Groq, DeepSeek
View on GitHub