Ingestion and capsule generation service for AgentNet that converts source content into structured records for downstream resolution.
Status: Alpha / Reference Implementation Docs: https://www.agent-net.ai Org: https://github.com/agentnet-ai
ant-capsulizer is the capsulizer component of the AgentNet reference architecture.
It is responsible for ingestion and capsule-oriented data transformation workflows that prepare source content for resolver consumption.
In a local development environment, it typically runs alongside ant-registrar for identity issuance and ant-resolver for downstream retrieval, with ant-orchestrator consuming resolved outputs.
Within the canonical AgentNet flow:
- Capsulizer
- Registrar
- Resolver
- Orchestrator
This repository provides the capsulizer role.
cp .env.example .env
npm installRun commands:
npm run devnpm startnpm run workernpm run schedulenpm run dev:seednpm run capsulize:repo
Runtime configuration is read from .env, with starter values in .env.example.
ANT_WORKER_OWNER_IDis required and must be a positive integer for seed/worker publish flows.- Discover it from registrar:
curl -s http://localhost:4002/v1/owners/ant-worker
Status: Alpha / Reference Implementation
These components are intended to demonstrate ANS-aligned architecture patterns.
Other core AgentNet reference components:
- https://github.com/agentnet-ai/AgentNet
- https://github.com/agentnet-ai/ant-capsulizer
- https://github.com/agentnet-ai/ant-registrar
- https://github.com/agentnet-ai/ant-resolver
- https://github.com/agentnet-ai/ant-orchestrator
Apache License 2.0. See LICENSE for details.