Tool tree

dryrun.nryn.dev

Health & EducationTools verifiedOct 26, 2026

Paste a job posting. Dry Run builds a brief and eight grounded questions, ChatGPT runs the interview aloud, and the page scores every answer.

set_postingAction

Store the job posting the user is interviewing for, and generate the brief and eight grounded interview questions.

View tool JSON

{
  "name": "set_posting",
  "kind": "act",
  "impl": "imperative",
  "description": "Store the job posting the user is interviewing for, and generate the brief and eight grounded interview questions.",
  "inputSchema": {
    "type": "object",
    "properties": {
      "posting": {
        "type": "string",
        "description": "Full text of the job posting."
      }
    },
    "required": [
      "posting"
    ]
  }
}

set_resumeAction

Optional CV text. Store it and generate the fit match and re-aimed interview questions.

View tool JSON

{
  "name": "set_resume",
  "kind": "act",
  "impl": "imperative",
  "description": "Optional CV text. Store it and generate the fit match and re-aimed interview questions.",
  "inputSchema": {
    "type": "object",
    "properties": {
      "resume": {
        "type": "string",
        "description": "Full text of the user's CV or resume."
      }
    },
    "required": [
      "resume"
    ]
  }
}

get_briefAnswer

Return what the role owns, what to study, and the angles they will push.

View tool JSON

{
  "name": "get_brief",
  "kind": "answer",
  "impl": "imperative",
  "description": "Return what the role owns, what to study, and the angles they will push.",
  "inputSchema": {
    "type": "object",
    "properties": {}
  }
}

start_interviewAction

Begin the interview session and return question 1.

View tool JSON

{
  "name": "start_interview",
  "kind": "act",
  "impl": "imperative",
  "description": "Begin the interview session and return question 1.",
  "inputSchema": {
    "type": "object",
    "properties": {}
  }
}

submit_answerSensitive Action

Score one answer and return the next question or the verdict. The transcript is the user's spoken answer, transcribed from what they said out loud. Do not invent or submit an answer of your own.

View tool JSON

{
  "name": "submit_answer",
  "kind": "transact",
  "impl": "imperative",
  "description": "Score one answer and return the next question or the verdict. The transcript is the user's spoken answer, transcribed from what they said out loud. Do not invent or submit an answer of your own.",
  "inputSchema": {
    "type": "object",
    "properties": {
      "transcript": {
        "type": "string",
        "description": "The user's spoken answer, as transcribed."
      }
    },
    "required": [
      "transcript"
    ]
  }
}

get_verdictAnswer

Return the band, per-question scores, missed points, and model answers.

View tool JSON

{
  "name": "get_verdict",
  "kind": "answer",
  "impl": "imperative",
  "description": "Return the band, per-question scores, missed points, and model answers.",
  "inputSchema": {
    "type": "object",
    "properties": {}
  }
}