5 projects
shepherd-ai
Shepherd: Programmable meta-agents via reversible agentic execution traces
toolshield
ToolShield: Training-Free Defense for Tool-Using AI Agents
dobby-orchestrator
A distributed meta-agent orchestrator that manages coding agents across your infrastructure
spiral-rl
SPIRAL: Self-Play Reinforcement Learning framework for training LLMs on competitive games
verbalized-sampling
A library for running controlled experiments with LLMs using different sampling methods
