AgentSight: System-wide AI agent profiling and monitoring with eBPF

# AgentSight: System-wide AI agent profiling and monitoring with eBPF

[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](https://opensource.org/licenses/MIT)
[![CI](https://github.com/eunomia-bpf/agentsight/actions/workflows/ci.yml/badge.svg)](https://github.com/eunomia-bpf/agentsight/actions/workflows/ci.yml)
[![arXiv:2508.02736](https://img.shields.io/badge/arXiv-2508.02736-b31b1b.svg)](https://arxiv.org/abs/2508.02736)
[![DOI:10.1145/3766882.3767169](https://img.shields.io/badge/DOI-10.1145%2F3766882.3767169-blue.svg)](https://dl.acm.org/doi/10.1145/3766882.3767169)

**English** | [中文](https://github.com/eunomia-bpf/agentsight/blob/master/README.zh-CN.md)

AI agents can run commands, rewrite files, spawn processes, and contact remote services—but when a run fails, stalls, or behaves unexpectedly, their logs rarely tell the whole story.

AgentSight is a local-first `top`/`strace`-like observability tool for AI agents. It connects prompts, model calls, and tool decisions to their real effects on your machine.

- **See what the agent is doing now.** Monitor active sessions, processes, resource, model and tool calls, file and network activity.
- **Understand failures and improve agent behavior.** Connect prompts and skills to errors, then turn them into better instructions and skills.
- **Find where time, tokens, and resources go.** Spot slow steps, retry loops, repeated model or tool calls, token-heavy sessions, and resource-hungry processes.
- **Audit data movement and security-sensitive effects.** See which services received requests, which files changed, and whether agents stayed and give safety/security suggestions.
- **Use it with your existing agents.** Run AgentSight around Claude Code, Codex, Gemini CLI, OpenCode, OpenClaw, or any command—without an SDK, proxy, or vendor integration.

No SDK, no proxy, no vendor integration. AgentSight observes with eBPF and TLS traffic tracing, so it works even when the agent is a
closed-source CLI. **✨ Zero SDK Required**

## Quick Start

```bash
cargo install agentsight
# or: wget https://github.com/eunomia-bpf/agentsight/releases/latest/download/agentsight && chmod +x agentsight
```


<div align="center">
  <img src="https://github.com/eunomia-bpf/agentsight/raw/master/docs/top-mode-demo.png" alt="AgentSight top live session view" width="1000">
  <p><em>Live sessions ranked by model, session tokens, health, process family, tool calls, file activity, and network activity</em></p>
</div>

```sh
agentsight top
```

<div align="center">
  <img src="https://github.com/eunomia-bpf/agentsight/raw/master/agentvis/examples/actplane-agent-nebula.gif" alt="Agent Nebula replay of Agent development across the ACTplane repository" width="1000">
  <p><em>Agent Nebula replays how coding agents read, write, create, rename, and delete files across the ACTplane repository</em></p>
</div>

```sh
agentsight vis
```

<div align="center">
  <img src="https://github.com/eunomia-bpf/agentsight/raw/master/docs/flamegraph-example/semantic-flamegraph-top200.svg" alt="Semantic flamegraph of the top 200 agent stacks" width="1000">
  <p><em>Width is system-effect weight; the uneven stack height shows prompt, tool-call, process, and effect paths ending at different depths. See the <a href="docs/agentpprof.md#example-flamegraphs">agentpprof guide</a> for the other profiles and how widths and stack depths are drawn.</em></p>
</div>

## 🚀 Why AgentSight?

### Traditional Observability vs. System-Level Monitoring

Application-level tools such as [LangSmith](https://docs.langchain.com/langsmith/observability-concepts), [Langfuse](https://langfuse.com/docs/observability/overview), and [Phoenix](https://arize.com/docs/phoenix/) are great for traces, prompts, tokens, evals, and latency when you own the application code. Gateway/proxy tools such as [Helicone](https://docs.helicone.ai/getting-started/integration-method/gateway) are useful when you can route provider traffic through a managed endpoint.

AgentSight focuses on the layer those tools often miss: what the agent actually does at the system boundary. It observes existing binaries and CLI agents without SDKs or proxies, then correlates LLM traffic with process execution, file access, and system activity.

| **Challenge** | **Application-Level Tools** | **AgentSight Solution** |
|---------------|----------------------------|------------------------|
| **Framework Adoption** | ❌ SDK, callback, or gateway integration per app | ✅ Drop-in system tracer, no code changes |
| **Closed-Source CLIs** | ❌ Limited to what the tool exposes or logs | ✅ Observes existing binaries and CLI agents from outside |
| **Agent-Controlled Logs** | ❌ Logs can be incomplete, disabled, or modified | ✅ Kernel-level events independent of app logging |
| **TLS LLM Traffic** | ❌ Visible when routed through SDKs/proxies | ✅ Captures plaintext at SSL/TLS calls without a proxy |
| **System Actions** | ❌ Often misses subprocesses and local file activity | ✅ T

... (truncated)