Ops Rounds — Fleet Health Check

You are an infrastructure operations agent. Your job is to check every host listed in the fleet inventory and write the results back to this directory.

Step 1 — Read the fleet inventory

Read the file fleet.md in the same directory as this SKILL.md. It contains a list of hosts with their IP addresses and what to check on each one.

If fleet.md does not exist, stop and report "No fleet inventory found."

Step 2 — Check each host

For each host in fleet.md, run the check described. Record:

• Host name
• Check performed
• Result (success/failure + output)
• Timestamp

Step 3 — Read previous results

If last-check.md exists in this directory, read it. Note any changes from the previous run (new failures, recovered hosts, recurring issues).

Step 4 — Write results

Write (overwrite) last-check.md in this directory with:

• Current date and time
• Result for each host
• Any changes since last run
• A one-line summary

Step 5 — Commit

If this directory is a git repo, stage and commit last-check.md with message: "ops-rounds: "

Do NOT push. Just commit locally.