nullboiler  by nullclaw

Orchestrates multi-step AI agent workflows

Created 2 months ago
315 stars

Top 85.6% on SourcePulse

GitHubView on GitHub
Project Summary

Summary

NullBoiler orchestrates complex AI agent workflows, defining execution graphs, managing shared state, and dispatching work to agents while tracking progress. It's designed for users building sophisticated AI systems who need a dedicated engine for coordinating distributed tasks, offering a modular approach that separates orchestration policy from task tracking and agent execution.

How It Works

This engine enforces strict architectural separation: a Tracker (e.g., nulltickets) holds task truth, the Orchestrator (nullboiler) applies scheduling and routing policies, and an Agent Runtime (e.g., nullclaw) executes tasks. NullBoiler's graph runtime supports nodes like task, agent, route, interrupt, send, transform, and subgraph, enabling features such as replay, checkpointing, and state injection for robust workflow management.

Quick Start & Requirements

The provided README does not detail a primary installation command (e.g., pip install) or specific non-default prerequisites like GPU, CUDA, or Python versions. Configuration relies on a JSON file (~/.nullboiler/config.json), with overrides possible via NULLBOILER_HOME or --config.

Highlighted Details

  • Features a rich workflow graph runtime with nodes for tasks, agents, routing, interrupts, sending, transformations, and subgraphs.
  • Supports advanced execution controls: replay, checkpoint forking, breakpoint interrupts, and post-start state injection.
  • Enables dynamic task dispatching with fan-out and configurable output shaping (output_key, output_mapping).
  • Provides template access to runtime states (state.*, input.*, item.*, config.*) and durable workflow memory via store.<namespace>.<key> and transform.store_updates, requiring a NullTickets base URL.

Maintenance & Community

No specific details regarding contributors, sponsorships, community channels, or roadmaps were found in the provided README text.

Licensing & Compatibility

The license type and compatibility notes for commercial use or closed-source linking are not explicitly mentioned in the provided README text.

Limitations & Caveats

NullBoiler's narrow focus means it requires integration with separate components like nulltickets (for tracking) and nullclaw or compatible agents (for execution) to form a complete system.

Health Check
Last Commit

1 week ago

Responsiveness

Inactive

Pull Requests (30d)
8
Issues (30d)
0
Star History
295 stars in the last 30 days

Explore Similar Projects

Starred by Gagan Bansal Gagan Bansal(Coauthor of AutoGen; Research Scientist at Microsoft Research), Elvis Saravia Elvis Saravia(Founder of DAIR.AI), and
1 more.

agent-framework by microsoft

1.8%
11k
AI agent and multi-agent workflow framework
Created 1 year ago
Updated 1 day ago
Feedback? Help us improve.