traceagent 0.1.1

TraceAgent — Local-first AI agent observability SDK. Trace LangChain, LangGraph, and any Python code. Open-source alternative to LangSmith.

TraceAgent

Local-first observability for AI agents. Open-source alternative to LangSmith. Trace LangChain, LangGraph, and any Python code with zero config.

Install

pip install traceagent

Quick Start

Option 1: Environment Variables (zero code changes, like LangSmith)

export TraceAgent_ENABLED=true
export TraceAgent_ENDPOINT=http://localhost:6832   # your TraceAgent server

import traceagent  # auto-traces all LangChain/LangGraph calls

result = graph.invoke({"messages": [HumanMessage(content="Hello!")]})
# traces appear at http://localhost:6832

Option 2: Explicit Callback

from traceagent import TraceAgent

lens = TraceAgent(server_url="http://localhost:6832")

result = graph.invoke(
    {"messages": [HumanMessage(content="Hello!")]},
    config={"callbacks": [lens.callback()]}
)

Option 3: Decorators & Context Managers

from traceagent import TraceAgent

lens = TraceAgent()

@lens.observe(name="search", kind="tool")
def search(query: str):
    return results

with lens.trace("My Pipeline") as t:
    result = search("hello")
    t.outputs = {"result": result}

Configuration

Env Variable	Description	Default
TraceAgent_ENABLED	Enable tracing	false
TraceAgent_ENDPOINT	Server URL	http://localhost:6832
TraceAgent_PROJECT	Project name	default
TraceAgent_TAGS	Comma-separated tags
TraceAgent_AUTO_TRACE	Auto-attach to LangChain	true when enabled

Server

The SDK sends traces to a TraceAgent server. Run it with Docker:

docker run -d -p 6832:6832 ashishgituser/TraceAgent-server:latest

Then open http://localhost:6832 for the dashboard.

See the full project at github.com/ashishgituser/TraceAgent.