BerriAI/lens

Última actividad de origen registrada: 8 de octubre de 2026 a las 23:55

Catálogo de SkillsMP actualizado: 12 de noviembre de 2026 a las 12:52

skills recopiladas: 1

Estrellas en GitHub: 21
Forks en GitHub: 7

Instalar con un asistente de IA

Copia este prompt en el asistente de IA que estés usando.

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/BerriAI/lens.

README del repositorio

Archivo de origen del README: README.md

Leer el README original

Mostrar README completo

<div align="center"> <a href="https://github.com/BerriAI/lens"> <picture> <source media="(prefers-color-scheme: dark)" srcset="assets/lens-logo-dark.png"> <source media="(prefers-color-scheme: light)" srcset="assets/lens-logo-light.png"> <img alt="LiteLLM Lens" src="assets/lens-logo-light.png" width="35%"> </picture> </a> <h3>Self-improving agents, on your infrastructure.</h3> **[Start locally](deploy/lens/README.md#start-locally)** with Git and Docker, or **[Set it up for me](docs/setup-with-agent.md)** with your coding agent <details> <summary>Copy a setup prompt</summary> ```text Set up Lens for this project using https://github.com/BerriAI/lens/blob/main/docs/setup-with-agent.md. Inspect the existing project and deployment first. Reuse a working Lens installation if one exists; otherwise start standalone Lens with ClickHouse using the documented source quickstart. Ask only for consequential missing choices. Preserve configuration, data, secrets, and my agent's model connection. Connect this project's instrumentation and verify a real run by its trace ID, including its input, output, and tool calls. Report the Lens URL, changes made, and verification results without exposing credentials. ``` </details> <br> <img src="assets/lens-hero.gif" alt="Agent swarms flow through the LiteLLM gateway into one trace per run, and Lens loops what it learns back into your agents" width="100%"> <br><br> **Your agents run. Lens watches every run, finds what went wrong, and feeds it back.**<br> **The next run is better.** <br> [Early access](https://forms.gle/3GC1Ner4vjthGWi18)   ·   [Docs](deploy/lens/README.md)   ·   [Contributing](CONTRIBUTING.md)   ·   [Launch post](https://docs.litellm.ai/blog/litellm-lens-launch)   ·   [4K film](assets/lens-hero-4k.mp4) <br><br> <img src="assets/lens-traces.png" alt="Lens traces" width="100%"> <br><br> <img src="assets/lens-finding.png" alt="Lens finding" width="100%"> <br><br> <img src="assets/lens-dataset.png" alt="Lens dataset" width="100%"> <br><br> <img src="assets/lens-mark.svg" alt="" width="72"> <br> **Run Lens on its own or open it inside LiteLLM.** Your traces stay in your own ClickHouse.<br> Point Claude Code or [Codex](docs/codex-cli.md) at them and let your agents improve your agents. <br> [Get started](deploy/lens/README.md) </div>

Cargando el README del repositorio…

Skills en este repositorio

clasificación pendiente

skill

ocupación

descripción

actualizado

url-state

sin clasificar

Read or write dashboard URL query state (tabs, filters, pagination, deep links, demo flags) with nuqs instead of raw history or URLSearchParams

Idioma del texto original: inglés

2026-10-08

Mostrando 1 de 1 skills recopiladas.