Back to directory

Agentry

An orchestration and organizational-memory layer for multi-agent systems, implementing the Agent Message Transfer Protocol for addressing and delivery between agents.

Agentry

Agentry provides universal agent@domain addressing with DNS-based discovery, a federated architecture for communication across organizational domains, and both pull-based inbox storage and push-based webhook delivery with at-least-once, idempotent guarantees. When an AI planner generates a dynamic workflow at runtime, Agentry executes it in parallel, sequential or conditional steps, injecting shared context across agents with token-aware selection and summarization, and discovering agents by capability rather than hardcoded addresses.

Its Context Graph acts as organizational memory, recording decision traces of what was decided, why, and what alternatives were rejected, so agents can query past precedent to inform current choices; traces are shared across agents and workflows and can be queried across federated organizational boundaries, preserving intent and reasoning across agent handoffs.

It is framework-agnostic, working with LangGraph, CrewAI or custom agents, and is built with Kubernetes-aware scheduling, AGNTCY schema validation, TLS 1.3 and API-key access control.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.