agentleFS
getedgehqgetedgehq/skills/domain-uptime-monitor/SKILL.md

domain-uptime-monitor

Skill5 starsChanged 15 days agoApache-2.0View on GitHub

Probe a list of domains with a real content check per domain, so a suspended deployment or a broken build that still returns HTTP 200 is caught, and alert only when a domain changes state. Use when asked to monitor uptime, watch production domains, set up a health check, or work out why a site looked up but was not.

  • Reads credentials

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

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

    Start from examples/domains.example.json.

    line 35
  2. Very specific:

    Read-only against your own domains, plus the post_check request where you configure one.

    line 99
  3. Very specific:

    Do not raise a domain to critical to make it more visible.

    line 103

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Domain Uptime Monitor

The failure this exists for

HTTP 200 is not proof a site is working:

  • A suspended or payment-disabled deployment serves a billing page with 200.
  • A broken build serves an empty shell with 200.
  • A partial deploy serves the old page while the API behind it is gone.
  • A misconfigured origin leaves the marketing page at 200 while every signed-in action fails.

A monitor that checks the status code reports all four as healthy. Every domain here declares what a working response actually contains.

Prerequisites

  • Python 3.9 or later, standard library only.
  • Outbound HTTPS to the domains you list.
  • No credential is needed to detect. Email alerting is optional and needs a Resend API key in RESEND_API_KEY and a sender domain verified on that account. Without the key, detection still runs and the report is still written; only the send is skipped. --no-email turns sending off explicitly.

notify_email has no default and will not get one. Ask the user for the address.

Writing the config

Start from examples/domains.example.json. Each domain takes a host, a tier, and a check.

Tier. critical domains alert and set the exit code. standard domains are probed and reported but never alert. Put revenue and API paths in critical and be strict about it: a monitor that alerts on everything gets muted, and then it alerts on nothing.

Check types.

typepasses whenuse for
statusstatus in allow, or any 2xx/3xx if allow is absentredirects, hosts where you have nothing better
markerstatus is expect_status (default 200) and marker is in the bodyany rendered page — this is the one that catches a 200 that is really off
jsonstatus matches, body parses, and the dotted field equals or contains the valuehealth endpoints; field may be nested, e.g. checks.db

Choosing a marker. Pick a string only the real page contains: a product name in the nav, a footer line, a specific heading. Do not use a string that a hosting error page would also contain, and do not use a build hash — it changes on every deploy and you will get a false alarm instead of a real one.

A host that is meant to return 401 or 403 is healthy when it does. Put those codes in allow, or it alerts every single day until someone mutes it.

Probing one host twice (the homepage and the checkout page, say) needs a distinct name on each entry. Results are keyed by that label, so the config is rejected rather than quietly reporting one fewer domain than you listed.

Compressed responses are decoded before the marker is looked for. Some servers gzip even when the request asks for identity, and a raw compressed body contains no readable marker at all. br cannot be decoded with the standard library and is left as-is, so a br-only origin needs a status or json check instead.

post_check probes an authenticated write path on a domain whose homepage is fine. Declare the signature of the break you have actually seen in fail_if (status, body_contains); anything in ok_status passes. Only add one where you know what the broken response looks like — a speculative write probe produces false alarms and sends traffic to an endpoint on every run.

Running it

python scripts/uptime_probe.py --config domains.json \
    --state .uptime/state.json --report out/report.md \
    --notify-email ops@yourdomain --email-from "Uptime <uptime@yourdomain>"

Exit codes: 0 all critical up, 1 a critical domain is down, 2 bad config, 3 the probe could not run. Put it on cron and let the exit code drive whatever you already have.

Alerting

Alerts fire on state change only, against --state. A domain that has been down for a week sends one email, not seven, and sends one more when it recovers.

A host with no previous state is new, not a transition, so adding a domain to the config never pages anyone.

A send failure is printed to stderr and never masks the detection result: the report and the exit code are correct whether or not the email went out.

Boundaries

  • Read-only against your own domains, plus the post_check request where you configure one.
  • RESEND_API_KEY is read from the environment inside the send function. Never pass it as an argument and never write it into the config file.
  • Do not raise a domain to critical to make it more visible. Tier is about who gets woken up.

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.