agentleFS
Sign inStart free
MicrosoftKnown publishermicrosoft/amplifier-smart-tool-reality-check/src/reality_check/skills/reaching-software/SKILL.md

reaching-software

Skill2 starsChanged 9 days agoMITView on GitHub

Reach software where it runs when checking it as a user: a local process, a deployed site, a Digital Twin Universe on this or another machine, or an app that is not running yet. Use when the app, the agent, and the user's view may be on different machines, or a URL does not answer.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Check reachability first with curl -sS -o /dev/null -w '%{http_code} %{time_total}\n' <url> from the place the user's view is.

    line 33
  2. Distinctive:

    Use only keys and hosts already configured.

    line 40

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

A check has three places, and each can be on a different machine: where you run, where the software runs, and where the user's view is (a browser, a screen, a terminal, a device). Find out which is which before you open anything.

Find where it runs

  • A URL in the request: open it from where the user's view will be, not only from here.
  • A port on this machine: ss -ltnp (Linux), lsof -iTCP -sTCP:LISTEN (macOS), or netstat -ano (Windows) shows what listens and on which address.
  • A universe: digital-twin-universe list and status --id <id> give its state and the URLs the host can open.
  • A repository and no running app: launch it in a new universe from a profile (digital-twin-universe create-profile, or launch with an existing one), as if deployed. Destroy it after the check unless asked to keep it.

localhost is relative

localhost means the machine or container the request runs in.

  • From inside a container or universe, the host's localhost is not reachable. Use the host's LAN address or name, or host.docker.internal where Docker provides it.
  • From a browser on another machine, use the serving machine's LAN IP, hostname, or Tailscale name, such as http://home-server-2:3000. A server bound to 127.0.0.1 answers only locally; say so in the report rather than changing how it is started without being asked.
  • Check reachability first with curl -sS -o /dev/null -w '%{http_code} %{time_total}\n' <url> from the place the user's view is. That shows the URL answers, not that the UI works.

Other machines

  • SSH: run commands there with ssh <host> '<command>', and forward a port with ssh -N -L 3000:localhost:3000 <host> so http://localhost:3000 here reaches the app there. Use only keys and hosts already configured; never create or change them.
  • Tailscale: tailscale status lists the machines on the tailnet and their names; a tailnet name works like a LAN name.
  • A universe on another machine: run digital-twin-universe there over SSH. Its URLs are localhost on that machine, so reach them through a port forward or that machine's name.

When it cannot be reached

Name what you tried, from where, and what answered: the URL, the command, and its output. A check that could not reach the software reports each item as not checked, never as a pass.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free