Devcon Archive logo
Devcon Devconnect Forum Blog
  • Watch
  • Event
    Event: background logo
    • Devcon 7
    • Devcon 6
    • Devcon 5
    • Devcon 4
    • Devcon 3
    • Devcon 2
    • Devcon 1
    • Devcon 0
    • Devconnect ARG
  • Categories
    Categories: background logo
    • Cryptoeconomics
    • Devcon
    • Developer Experience
    • Coordination
    • Core Protocol
    • Layer 2s
    • Real World Ethereum
    • Cypherpunk & Privacy
    • Security
    • Applied Cryptography
    • Usability
  • Playlists

Suggested

Loading results..

View all

About Devcon —

Devcon is the Ethereum conference for developers, researchers, thinkers, and makers.

An intensive introduction for new Ethereum explorers, a global family reunion for those already a part of our ecosystem, and a source of energy and creativity for all.

  • Watch
  • Devcon
  • Devconnect
  • Forum
  • Blog

Get in touch

devcon@ethereum.org

Subscribe to our newsletter

Crafted with passion ❤️ at the Ethereum Foundation

© 2026 — Ethereum Foundation. All Rights Reserved.

devcon8 / why your observability stack is making your agents stupid and ours isnt anymore

  • YouTube
Why your observability stack is making your agents stupid (and ours isn't anymore) preview
  • Details

Why your observability stack is making your agents stupid (and ours isn't anymore)

Duration: NaN:NaN:NaN

Speaker: Stefan Starflinger

Type: Talk

Expertise: No prior knowledge

Event: Devcon

Date: Nov 2026

The 2022-era observability setup is friction for humans and poison for agents: Grafana in front of ClickHouse, Prometheus, and Loki, three query languages, and auth in a browser. We did the obvious integration first, building one MCP tool per datasource. It failed instructively: a single query would dump ~7000 rows of JSON into the context window, rotting the agent's context until it forgot its original task. We'll walk through the rebuild: Panda, a CLI and MCP server sharing one backend, with a sandboxed Python runtime where raw rows live and die and only summaries return to the model. Humans and agents use the same surface. We'll cover the zero-credential sandbox that makes LLM code against prod safe, and the semantic search that solves discovery. Any core dev can use it today via panda init. You'll leave with a pattern to steal: give agents a computer, not a pile of tools.

Categories

Core ProtocolAIOpen Source