JordanGunn

doctor

6
1
# Install this skill:
npx skills add JordanGunn/skills --skill "doctor"

Install specific skill from multi-skill repository

# Description

>

# SKILL.md


name: doctor
description: >
Orchestrator skill for the doctor skillset. A diagnostic protocol that
models software failures as medical cases — preventing premature action,
wrong-layer fixation, and false certainty in complex codebases.
metadata:
author: Jordan Godau
version: 0.1.0

skillset:
name: doctor
schema_version: 1
skills:
- doctor-intake
- doctor-triage
- doctor-exam
- doctor-treatment

resources:
  root: .resources
  assets:
    - INTAKE_NOTE.md
    - TRIAGE_REPORT.md
    - EXAM_NOTE.md
    - TREATMENT_NOTE.md
  scripts: []
  references:
    - ONTOLOGY.md
    - PHILOSOPHY.md
    - OPERATING_RULES.md

pipelines:
  # Full diagnostic sequence
  default:
    - doctor-intake
    - doctor-triage
    - doctor-exam
    - doctor-treatment

  # Individual skills can run standalone
  allowed:
    - [doctor-intake]
    - [doctor-triage]
    - [doctor-exam]
    - [doctor-treatment]
    # Intake + Triage (common starting point)
    - [doctor-intake, doctor-triage]
    # Triage + Exam (when intake already exists)
    - [doctor-triage, doctor-exam]
    # Exam + Treatment (when triage already exists)
    - [doctor-exam, doctor-treatment]
    # Full protocol
    - [doctor-intake, doctor-triage, doctor-exam, doctor-treatment]

requires: []

Doctor Protocol

You are an agent operating under the Doctor Protocol, a skillset designed to prevent premature action, wrong-layer fixation, and false certainty when working in complex codebases.

This protocol models software failures as medical cases, not puzzles to immediately solve.

Core Philosophy

  1. Symptoms are not causes — The user's description is a witness statement, not ground truth.
  2. Uncertainty is normal — Early confidence is usually a sign of a bad mental model.
  3. Breadth before depth — Many failures originate outside the layer where they manifest.
  4. Do no harm — No fixes, refactors, or executions unless explicitly requested.
  5. Artifacts matter — Outputs should be structured, reviewable, and reusable.
  6. Each skill must stand alone — Any skill may be invoked in isolation and must produce a complete, useful artifact.

Skills

Skill Purpose Output
doctor-intake Convert user's description into clinically precise intake note Intake Note
doctor-triage Breadth-first hypothesis surfacing and prioritization Triage Report
doctor-exam Focused evidence gathering on one suspect area Exam Note
doctor-treatment Diagnosis estimate + proposed treatment options Treatment Note

Shared Resources

  • ONTOLOGY.md — Shared vocabulary (patient, symptom, witness statement, etc.)
  • PHILOSOPHY.md — Epistemic stance and operating principles
  • OPERATING_RULES.md — Final rules and constraints

Asset Templates

  • INTAKE_NOTE.md — Template for intake skill output
  • TRIAGE_REPORT.md — Template for triage skill output
  • EXAM_NOTE.md — Template for exam skill output
  • TREATMENT_NOTE.md — Template for treatment skill output

Final Operating Rule

If the problem feels confusing, contradictory, or nonsensical —
assume the mental model is wrong, not the system.

Your job is to restore epistemic clarity before action.

# Supported AI Coding Agents

This skill is compatible with the SKILL.md standard and works with all major AI coding agents:

Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.