The Extensible, Multi-Agent Personal AI Sidekick
Manager-agent pattern for orchestrating OS+browser+code agents, but Cursor and Continue already own this space.
AgentLog is a lightweight event bus for AI agents built on append-only JSONL logs. Agents publish and subscribe to events via HTTP/SSE, enabling simple, replayable, and debuggable multi-agent workflows.
JSONL files you can tail -f beats complex Kafka setup for agent event streams.
Developers building multi-agent systems
Kafka · NATS · Redis Streams
I built a small project called AgentLog.
The core idea is very simple, topics are just append-only JSONL files.
Agents publish events over HTTP and subscribe to streams using SSE.
The system is intentionally single-node and minimal for now.
Future ideas I’m exploring: - replayable agent workflows - tracing reasoning across agents - visualizing event timelines - distributed/federated agent logs
Curious if others building agent systems have run into similar needs.
Manager-agent pattern for orchestrating OS+browser+code agents, but Cursor and Continue already own this space.
Recovers wiped projects from Claude Code logs when git history won't help.
Stateless agents with S3 storage cuts Kafka+Flink costs by 92%, but confluent-kafka wire protocol still needs fixes.
Applies MIT CSAIL research patterns to make AI agent decisions auditable and traceable.
Time-travel SQL on append-only storage, but PostgreSQL temporal tables already exist.
Task dependency graphs in JSONL beat markdown blobs for agent-driven planning.