Skip to content
TakoAPI
Back to skills

patiently-ai

by nickjlamb

CLI Utilities
475Downloads reported by clawskills.sh
0Views
0 likesSign in to like this skill
—Not yet rated
Your rating

Brief

Patiently AI simplifies medical documents for patients.

Description

What This Skill Does

Patiently AI takes medical documents and re-explains them in plain, personalised language. It accepts doctor's letters, lab results, prescriptions, discharge summaries, photos, PDFs, and audio recordings of consultations. Output is structured with a summary, section-by-section breakdown, glossary of terms, and suggested questions for the patient's healthcare provider.

Patients often leave appointments with documents they cannot parse, and re-contacting a clinic for clarification takes time, so this provides an immediate plain-language breakdown without requiring medical knowledge.

When to Use It

  • Understanding what a blood test result means before a follow-up appointment
  • Helping an elderly parent make sense of a hospital discharge letter
  • Reading a specialist's clinic note after a referral
  • Decoding a prescription and its listed side effects
  • Preparing questions to ask a doctor at the next visit

Example Workflow

INPUT

User asks: 'Can you explain this blood test?' and attaches an image of their lab report

AGENT
  1. 1

    Ask the user for their preferred reading level, tone, and output length, or apply defaults

  2. 2

    Extract values and reference ranges from the attached image

  3. 3

    Summarise what the test measured in 2-3 plain-language sentences

  4. 4

    Break down each result, noting whether it falls within the normal range

  5. 5

    Define medical terms used and suggest 3-5 follow-up questions for the patient's doctor

OUTPUT

A structured explanation covering the test summary, per-result breakdown, a glossary of terms, and a list of questions the patient can bring to their next appointment

Installation

clawhub install nickjlamb/patiently-ai

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links