dug.sh

dug_investigate

Answer a question that takes more than one lookup, and build the case on the page while you do. You choose the commands and the order: this is your plan, not a preset. Each screen is left on the page, in sequence, under the question that produced it, so the person watching ends up with the evidence, not your summary of it. Prefer this over calling the single-command tools yourself whenever the person has described a symptom instead of naming a lookup, and whenever one answer won’t settle it. Steps use the same grammar as the other tools: a command name, {target} where the domain goes, and an optional second argument; each step runs once per target, in order. For example "why is mail from this domain going to spam" with steps ["MAIL {target}","SPF {target}","DIG {target} MX","RDAP {target}"], or "has my dns change taken effect" with steps ["PROP {target}","TTL {target}","NS {target}","DNSSEC {target}"]. A step naming a command that doesn’t exist is skipped and reported back to you; the rest still run.

Page: /

Input schema

{
  "type": "object",
  "properties": {
    "question": {
      "type": "string",
      "description": "The question being answered, in the words the person would use. It becomes the heading the evidence is filed under. Not a restatement of the commands."
    },
    "targets": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "The domains or hosts under investigation, bare names, one entry when there is one. The steps run against each in this order."
    },
    "steps": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "The commands to run, in order, each a full command line such as \"MAIL {target}\" or \"DIG {target} MX\". Three to five is usually right: enough to be conclusive, few enough to read."
    }
  },
  "required": [
    "question",
    "targets",
    "steps"
  ]
}