Nine tools, one human gate. The agent can do everything except decide.

Tells a practice whether a named AI vendor will sign a business associate agreement on a given subscription tier, and what the eligible path looks like if it will not. Use it the moment someone names a tool they already use, even before any readiness check has started. Returns the vendor verdict, the consumer-versus-enterprise distinction and a line to verify with the vendor. This is general education and never a determination about the practice.

{
  "type": "object",
  "properties": {
    "vendor": {
      "type": "string",
      "enum": [
        "openai_chatgpt",
        "anthropic_claude",
        "google_gemini",
        "microsoft_copilot",
        "nuance_dax",
        "abridge",
        "suki",
        "freed",
        "nabla",
        "other"
      ]
    },
    "tier": {
      "type": "string",
      "enum": [
        "consumer_free",
        "consumer_paid",
        "team_business",
        "enterprise",
        "api",
        "not_sure"
      ]
    }
  },
  "required": [
    "vendor"
  ],
  "additionalProperties": false
}

Explains the recording-consent rule in one US state and what it means for an ambient scribe sitting in the room. Use it when a practice names its state or asks whether it needs the patient's agreement before recording. Returns whether the state is all-party or one-party, any AI-disclosure statute we are confident about, and the practical implication. General education only, not a legal determination.

{
  "type": "object",
  "properties": {
    "state": {
      "type": "string",
      "enum": [
        "AL",
        "AK",
        "AZ",
        "AR",
        "CA",
        "CO",
        "CT",
        "DE",
        "DC",
        "FL",
        "GA",
        "HI",
        "ID",
        "IL",
        "IN",
        "IA",
        "KS",
        "KY",
        "LA",
        "ME",
        "MD",
        "MA",
        "MI",
        "MN",
        "MS",
        "MO",
        "MT",
        "NE",
        "NV",
        "NH",
        "NJ",
        "NM",
        "NY",
        "NC",
        "ND",
        "OH",
        "OK",
        "OR",
        "PA",
        "RI",
        "SC",
        "SD",
        "TN",
        "TX",
        "UT",
        "VT",
        "VA",
        "WA",
        "WV",
        "WI",
        "WY"
      ]
    }
  },
  "required": [
    "state"
  ],
  "additionalProperties": false
}

Opens the ten-question readiness check on screen with the practice profile filled in. Use it once, after you know the specialty, how many clinicians there are, the state and which AI tools are already in use. Returns the list of question ids to work through in order. Nothing here is a determination about the practice, and no patient information should ever be entered.

{
  "type": "object",
  "properties": {
    "specialty": {
      "type": "string",
      "enum": [
        "primary_care",
        "therapy_behavioral",
        "dental",
        "dermatology",
        "pediatrics",
        "other"
      ]
    },
    "clinician_count": {
      "type": "integer",
      "minimum": 1,
      "maximum": 10
    },
    "state": {
      "type": "string",
      "enum": [
        "AL",
        "AK",
        "AZ",
        "AR",
        "CA",
        "CO",
        "CT",
        "DE",
        "DC",
        "FL",
        "GA",
        "HI",
        "ID",
        "IL",
        "IN",
        "IA",
        "KS",
        "KY",
        "LA",
        "ME",
        "MD",
        "MA",
        "MI",
        "MN",
        "MS",
        "MO",
        "MT",
        "NE",
        "NV",
        "NH",
        "NJ",
        "NM",
        "NY",
        "NC",
        "ND",
        "OH",
        "OK",
        "OR",
        "PA",
        "RI",
        "SC",
        "SD",
        "TN",
        "TX",
        "UT",
        "VT",
        "VA",
        "WA",
        "WV",
        "WI",
        "WY"
      ]
    },
    "ai_tools": {
      "type": "array",
      "items": {
        "type": "string",
        "enum": [
          "ambient_scribe",
          "general_chatbot",
          "ehr_builtin_ai",
          "dictation",
          "other",
          "not_sure"
        ]
      },
      "minItems": 1,
      "maxItems": 6
    }
  },
  "required": [
    "specialty",
    "clinician_count",
    "state",
    "ai_tools"
  ],
  "additionalProperties": false
}

Records the practice's answer to one of the ten questions and advances the visible progress. Use it once per question, after the human has actually answered in their own words, choosing the option that matches. Returns the recorded answer and which question comes next. Answers stay in the browser; never pass anything that identifies a patient.

{
  "type": "object",
  "properties": {
    "question_id": {
      "type": "string",
      "enum": [
        "q01",
        "q02",
        "q03",
        "q04",
        "q05",
        "q06",
        "q07",
        "q08",
        "q09",
        "q10"
      ]
    },
    "answer": {
      "type": "string",
      "description": "One option value for that question. For q02 pass a comma-separated list of tool values."
    }
  },
  "required": [
    "question_id",
    "answer"
  ],
  "additionalProperties": false
}

Reveals the readiness score out of 100 and the three sharpest exposures on screen, with the animated score ring. It is always free and never gated behind an email. Use it as soon as the questions are answered, or at any point to show where the practice stands. Returns the score, the ranked risks and the plain-language reason each one matters.

{
  "type": "object",
  "properties": {},
  "required": [],
  "additionalProperties": false
}

Explains one exposure in plain language: what is exposed, why it matters in one sentence, what fixes it, and the statute reference behind it. Use it when the practice asks why a particular gap is on their list. Returns the explanation and the related safeguards from the thirty-control library. This is general education, not a determination about the practice.

{
  "type": "object",
  "properties": {
    "risk_id": {
      "type": "string",
      "enum": [
        "risk_vendor_baa",
        "risk_consumer_tier",
        "risk_recording_consent",
        "risk_human_review",
        "risk_risk_analysis",
        "risk_written_rule",
        "risk_officer_training"
      ]
    }
  },
  "required": [
    "risk_id"
  ],
  "additionalProperties": false
}

Offers the personalised remediation roadmap. This tool takes no arguments and cannot accept an email address: it opens a dialog on screen where the human types their own address and ticks consent themselves. Use it after the score has been revealed and the practice wants the full roadmap. It waits for the human, then returns whether they sent it and opens the roadmap page.

{
  "type": "object",
  "properties": {},
  "required": [],
  "additionalProperties": false
}

Hands the practice over to the 21-day adoption mission on lab.hipaa.technology, carrying their profile and top exposure in the link. Use it once the roadmap is on screen and they want a paced way through it. Opens the link in a new tab and shows it on screen. Returns the link that was opened.

{
  "type": "object",
  "properties": {},
  "required": [],
  "additionalProperties": false
}

Shows the booking card for a 20-minute review with a human from Safe AI Technology. Use it when the practice asks to talk to a person rather than work through the roadmap alone. It only puts the card and the link on screen — the human books the time themselves. Returns the booking link.

{
  "type": "object",
  "properties": {},
  "required": [],
  "additionalProperties": false
}