Skip to content

marcusschiesser/agentpond

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

130 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AgentPond — private trace storage and local analytics for AI agents

Store agent traces remotely. Analyze them locally. Keep control of the data.

CI status npm version npm downloads MIT license Node.js version

AgentPond is a lightweight trace backend and CLI for AI applications. It keeps raw traces in object storage you control and syncs them into a local DuckDB cache for fast analysis by your coding agent. It supports all bucket providers from Files SDK such as Amazon S3, Google Cloud Storage, R2, and MinIO; For platforms like Firebase, Supabase, and Vercel there are explicit one-command quick starts.

How it works

AgentPond data flow from agent traces through object storage and local CLI analysis

Object storage is the durable source of truth. The local DuckDB database is a rebuildable analytical cache, so production traces stay in your infrastructure without requiring an always-on analytics database.

Getting started

Start with the Manual deployment setup to configure object storage, instrument your application, and sync its traces into AgentPond.

For Firebase, Supabase, and Vercel projects, to get started, just run:

npx agentpond init

The command installs AgentPond's instrumentation and analytics skills, and prints a install prompt for your coding agent. Your coding agent inspects then the application, proposes a setup, implements it after confirmation, and verifies a real trace end to end.

Once the application has emitted a trace:

npx agentpond sync
npx agentpond traces list --limit 10

Features

  • Direct OpenTelemetry export through Files SDK, Firebase Storage, Supabase Storage, and Vercel Blob
  • Langfuse-compatible and OTLP HTTP ingestion adapters
  • Incremental object-store synchronization
  • Local DuckDB projections for traces, observations, sessions, and scores
  • Focused CLI commands plus arbitrary SQL
  • Human annotations represented as scores
  • Agent skills for instrumentation and trace investigation

Intentional non-goals

AgentPond does not provide a web UI, hosted trace storage, prompt management, datasets, or always-on Postgres, ClickHouse, Redis, and worker infrastructure. Use a full observability platform when shared dashboards or non-technical workflows are required.

Documentation

Development

pnpm install
pnpm cli --help
pnpm test

About

Store agent traces remotely. Analyze them locally. Keep control of the data.

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages