evilmartians/agent-prism

React components for visualizing traces from AI agents

What it solves

AgentPrism solves the problem of debugging AI agents by transforming complex, JSON-heavy trace data into clear, visual diagrams. It prevents critical information about an agent's behavior—such as plans, actions, and retries—from being lost in raw logs.

How it works

It is a library of React components that visualizes traces from AI agents. It uses adapters to normalize data from sources like OpenTelemetry (OTLP) and Langfuse into a format optimized for UI rendering. These traces are then displayed in a hierarchical timeline, allowing users to see LLM calls, tool executions, and agent workflows.

Who it’s for

Developers building AI agents who need a visual way to inspect and debug their agentic workflows and trace data.

Highlights

  • Comprehensive Visualization: Includes a TraceViewer for a complete interface, or individual components like TraceList, TreeView, and DetailsView for custom layouts.
  • Broad Data Support: Provides built-in adapters for OpenTelemetry and Langfuse formats.
  • Semantic Convention Support: Recognizes standard GenAI attributes (e.g., gen_ai.*, llm.*, retrieval.*) to display model, tokens, and costs.
  • Customizable Theming: Uses semantic tokens and Oklch colors for flexible styling via Tailwind CSS.

Related

  • Project
    ServiceNow/AgentLab

    A framework for developing and evaluating web agents across various benchmarks, featuring large-scale parallel experimentation and trace visualization tools.

  • Project
    liveloveapp/hashbrown

    An open-source framework for building AI agents that run in the browser, enabling developers to easily embed intelligence into React and Angular components.

  • Project
    rajudandigam/agent-inspect

    A local-first observability and debugging tool for TypeScript AI agents that turns execution traces into readable trees and deterministic trajectory checks for CI.

  • Project
    canwhite/AgentEval

    A transparent HTTP proxy that captures Agent-LLM traffic to provide automated grading, behavioral diagnosis, and source-code probing for AI agent evaluation.

  • Project
    nullpointexception-i/agent-sphere

    AgentSphere is a Java‑Spring/React platform that runs LLM‑driven ReAct loops, supports multi‑agent sub‑runs, real browser automation via a Chrome extension, model routing, persistent multi‑level memory, human‑in‑the‑loop clarification, OIDC SSO, and an embeddable chat widget.