AgenticLens

{ width="420" }

Open-source evaluation and profiling for production-ready agentic AI systems.

AgenticLens helps teams profile LLM applications, measure cost and latency, and turn traces into practical recommendations for production agent systems.

What It Measures

Dimension	What AgenticLens captures
Cost	Prompt tokens, completion tokens, provider pricing, monthly projections
Latency	Step-level runtime and tokens per second
Workflow shape	Planner, retriever, tool, memory, and final-response steps
Waste patterns	Repeated prompts, excessive chunks, duplicate tool calls, long history
Quality risk	Confidence and risk notes for optimization recommendations
Resilience	Fault-injection outcomes through the chaos_events schema extension

Why It Matters

Production agent systems fail in ways that ordinary request logs rarely explain. Token cost can drift across memory, retrieval, planning, and tool use. Reliability can degrade silently when an upstream tool fails. AgenticLens keeps these signals local, inspectable, and exportable so teams can compare workflows across versions.

Documentation

• Workflow schema specification
• Export formats
• RAG chunk utility
• Technical specification

Quickstart

from agenticlens import profile, step

with profile("Customer Support"):
    with step("Planner", type="planner") as s:
        response = planner_llm.invoke(prompt)
        s.record(response)

uv run agenticlens profile examples/recommendations_demo.py --save workflow.json
uv run agenticlens analyze workflow.json

Current Status

AgenticLens is early-stage open-source software. The core profiler, CLI, export formats, pricing model, and recommendation engine are implemented. Integrations for agent frameworks, trace formats, and workflow explorers are active roadmap priorities.