beads

Manage Git-backed issues with CLI commands for dependencies and memory.

Updated Jan 21, 2026
One-click install
npx skills add https://github.com/mbarnes-code/multi-agent-vllm --skill beads-mbarnes-code
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: beads
Source: https://github.com/mbarnes-code/multi-agent-vllm/tree/main/features/beads/claude-plugin/skills/beads
Command: npx skills add https://github.com/mbarnes-code/multi-agent-vllm --skill beads-mbarnes-code

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a robust, Git-backed issue tracking system that persists across conversation compactions and long periods of inactivity, ensuring that complex multi-session work and critical project memory are never lost.

Core Features & Use Cases

  • Persistent Memory: Tracks issues, dependencies, and notes that survive AI context window limitations and session resets.
  • Dependency Management: Organizes tasks with explicit blockers, related items, and hierarchical structures.
  • Multi-Session Work: Ideal for tasks spanning days or weeks, allowing agents to resume work with full context.
  • Use Case: An AI agent working on a complex software feature can use beads to track sub-tasks, dependencies, and design decisions. Even if the conversation is compacted or the agent restarts days later, it can resume exactly where it left off by consulting the beads database.

Quick Start

Use the beads skill to create a new task named 'Implement user authentication flow'.

Frequently Asked Questions about beads

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How does persistent task memory work for AI agents across conversation compactions?

Persistent task memory for AI agents uses a Git-backed issue tracking database to store tasks, dependencies, and notes. This ensures project context survives context window limitations and session resets, allowing agents to resume multi-session work exactly where they left off.

How do I track complex issue dependencies and blockers for multi-session AI workflows?

To track complex issue dependencies in AI workflows, use CLI commands for issue creation, status updates, and dependency linking. This organizes tasks with explicit blockers, related items, and hierarchical structures that persist indefinitely across sessions.

Do I need Git integration to use persistent issue tracking for AI agents?

Yes, Git integration is required because the issue tracking system is Git-backed to ensure persistent memory and context survival across AI conversation compactions. You also need bd CLI v0.47.0 or higher installed for full functionality.

What is the best way to maintain project context when an AI agent restarts after days of inactivity?

The best way to maintain project context after AI agent inactivity is consulting a Git-backed project memory database. By tracking sub-tasks, dependencies, and design decisions there, the agent can resume work with full context even after days or weeks.

Are there limitations when using Git-backed issue tracking for AI context survival?

A limitation of Git-backed issue tracking for AI context survival is the strict dependency on bd CLI v0.47.0 or higher for full functionality. Without this specific CLI version installed, the persistent memory and dependency management features cannot operate.