a PrompTugal.news site · the agentic team · the operating model 1/7
Lisbon Tech Fringe · Dinis Cruz with Angelo · 11 Nov 2026

The execution ladder

"What they are depends on where they are and what they can do."
E0

compute

raw code execution: a script, a build, a CI job
admin/build/build.mjs in GitHub Actions

GitHub ActionsAn AWS Lambda with a KMS signing key
E1

model call

a prompt in, an answer out (the first chatbots)
the second opinion via the router

A model call via OpenRouterA chat sessionA Cloudflare Worker with Workers AI
E2

loop

a simple agentic workflow: the model's output feeds its next step, without tools
a chain of calls in one script

E3

loop with tools

the loop can act: shell, files, web, git
a Claude Code session; a ChatGPT agent

A ChatGPT agentAn AWS Bedrock AgentCore agent
E4

schedules

it can wake itself later (create and keep schedules)
every agent's hourly wake

A Claude Code cloud agentA laptop agent with OllamaA Cloudflare Agents SDK agent
E5

spawn

it can start new agents, sessions and environments: the control plane
the lead starting architect.1

The lead

Tools

cut across the rungs: a loop acts through them (E3), and a single call can be handed them too.

Storage

what it can mount: a file system, git, vaults. State in, evidence out.

Identity

a key on every rung, even a chat. An ephemeral key proves the agent existed at that moment.

lisbontechfringe.com · the agentic teamdata/operating-model.json · work/48 · a draft for Dinis and Angelo to correct

Lisbon Tech Fringe is independent of Web Summit and not run by its organisers. For anything official, see the organisers' official site.