Skip to content

Latest commit

 

History

History
52 lines (38 loc) · 2.61 KB

File metadata and controls

52 lines (38 loc) · 2.61 KB

Agent Guide

This repository contains independent example applications for the Green Metrics Tool. Most agent work should start from a scenario directory that contains one or more usage_scenario*.yml files.

How To Navigate

  • Start with scenario-index.yml to find the right example quickly.
  • Then read the nearest AGENTS.md in the target directory.
  • Use usage_scenario*.yml as the source of truth for services, commands, and flow order.
  • Use compose.yml and Dockerfile* files for runtime and build details.
  • Treat README.md as human-oriented background unless the local AGENTS.md tells you otherwise.

Repository Conventions

  • Each scenario directory is intended to be runnable on its own.
  • Two nested families contain child examples instead of a single root scenario:
    • green-software-foundation-sci/*
    • wordpress-vs-hugo/*
  • .skiptest means the example is excluded from automated repo testing. Do not assume the scenario is broken.
  • If a scenario and a README disagree, prefer the scenario file.

Important Caveats

  • GUI/X11 examples such as firefox-X11 and blauer-engel-okular-xdotool depend on host display state and usually require unsafe volume mounts.
  • ai-model/usage_scenario_gpu.yml requires an NVIDIA-capable Docker setup and unsafe mounts.
  • Seeded CMS examples such as drupal, wordpress-official-data, and wordpress-vs-hugo/wordpress-minimal need database startup time before browser or curl steps run.
  • long-test intentionally runs for roughly three hours.
  • network-test depends on external network downloads, so failures may be environmental rather than code regressions.

Scope Boundaries

  • Ignore vendored payload content under paths such as html/, wp-content/, and similar application data trees unless the task is explicitly about the bundled sample app itself.
  • Nested third-party README.md files inside application payloads are not repository navigation docs.

Recommended Agent Workflow

  1. Pick a scenario from scenario-index.yml.
  2. Read the local AGENTS.md.
  3. Inspect the referenced usage_scenario*.yml.
  4. Open compose.yml, Dockerfile*, and helper scripts named in the scenario.
  5. Only descend into bundled app content if the scenario or task requires it.

Running an example scenario

If you want to check an example application if it is working run this:

cd /home/node/green-metrics-tool
python3 runner.py --uri /home/node/example-applications --filename stress/usage_scenario.yml` --dev-no-sleeps --dev-no-system-checks

You must replace stress/usage_scenario.yml with the path to the scenario you want to run.