domains / documents / treatment-plan-documentation / treatment-plans

treatment-plans

Skill card

combined

The copies define two different procedures for the same skill name. K-Dense-AI/scientific-agent-skills treats treatment-plans as a documentation-only formatting/validation workflow that explicitly forbids clinical decision-making, uses local JSON templates, and requires a DRAFT notice and release gate. The other three copies (Vibe-Skills, medical-research-skills, OpenClaw-Medical-Skills) treat it as a LaTeX/PDF generation workflow that creates concise clinical treatment plans, includes SMART goals, evidence-based interventions, and mandatory AI-generated schematics. These are not shortenings or wrappers of one another; they disagree on purpose, output format, safety scope, and required steps, so crowning one would hide a real alternative.

Repositories

Other copies

Combined skill

# Treatment-Plan Documentation

## Purpose

This skill covers two distinct workflows for treatment-plan documentation. Select the appropriate workflow based on the task:

- **Documentation-Only Workflow** (K-Dense-AI/scientific-agent-skills): Formats and structurally validates local treatment-plan documentation after clinical decisions have already been supplied and verified by authorized licensed professionals. Does not perform clinical decision-making.
- **LaTeX/PDF Generation Workflow** (Vibe-Skills, medical-research-skills, OpenClaw-Medical-Skills): Generates concise, focused medical treatment plans in LaTeX/PDF format for all clinical specialties, including SMART goals, evidence-based interventions, and professional formatting.

## Conflicts

- **K-Dense-AI/scientific-agent-skills** forbids clinical decision-making, therapy selection, dose selection, and outcome prediction. The other three copies instruct the agent to generate evidence-based interventions and clinical recommendations.
- **K-Dense-AI/scientific-agent-skills** outputs local JSON packages and requires a DRAFT notice and release gate. The other three copies output LaTeX/PDF documents.
- **K-Dense-AI/scientific-agent-skills** requires no network or third-party packages. The other three copies require TeX Live and specific LaTeX packages.
- **Vibe-Skills and OpenClaw-Medical-Skills** make at least one AI-generated schematic mandatory. **medical-research-skills** lists it as optional. **K-Dense-AI/scientific-agent-skills** does not require figures.

## Documentation-Only Workflow

### Hard Safety Boundary

This workflow only formats and validates documentation of decisions already made, supplied, and verified by authorized licensed professionals. Never use it to diagnose, select therapies, choose medications or doses, start/stop/titrate treatments, check interactions, infer missing clinical content, triage, predict outcomes, or replace clinical review.

### Required Visible Notice

Every component and derived schedule must display:

> **DRAFT — NOT MEDICAL ADVICE — DOCUMENTATION-ONLY — AUTHORIZED CLINICIAN SIGN-OFF REQUIRED**

### Data Gate

Prefer synthetic or qualified de-identified structured manifests. Do not place patient names, medical-record numbers, contact details, dates of birth, addresses, free-text notes, images, or other direct identifiers in examples. For real-patient data, work only in a locally authorized environment, use minimum necessary information, do not send content to external tools, and require qualified privacy review.

### Allowed Inputs

Accept only bounded UTF-8 JSON objects built from the six generic templates: source_fact_manifest, clinician_authored_intervention, goals_monitoring_checkpoint, informed_preference_shared_decision, transition_reconciliation, and intended_use_handoff.

### Workflow

1. Establish authority and intended use: confirm the accountable clinical owner, authorized signatory, jurisdiction, institution, setting, document owner, local policies, retention rule, and intended recipients. Keep the release gate blocked until every required review is complete.
2. Generate a generic package using the bundled Python script. The generator copies all six templates and does not create clinical content.
3. Transcribe supplied decisions without inference: copy only clinician-authored facts and interventions from verified local sources. Preserve source locators, versions, author role, verification role, and verification time. Leave missing fields unresolved.
4. Validate structure: run bundled validation scripts against local paths. Reports identify rule codes and field paths, not clinical values.
5. Release gate: only the authorized local workflow may set the release gate. Structural success never removes the DRAFT notice.

## LaTeX/PDF Generation Workflow

### Overview

Generate concise, focused treatment plans (1–4 pages standard) across all medical specialties with full regulatory compliance. Prioritize brevity and clinical actionability.

### Critical Principles

1. **CONCISE & ACTIONABLE**: Treatment plans default to 3–4 pages maximum, focusing only on clinically essential information.
2. **Patient-Centered**: Plans must be evidence-based, measurable, and compliant with healthcare regulations.
3. **Minimal Citations**: Use brief in-text citations only when needed; avoid extensive bibliographies.

### When to Use

- Creating individualized treatment plans for patient care
- Documenting therapeutic interventions for chronic disease management
- Developing rehabilitation programs
- Writing mental health and psychiatric treatment plans
- Planning perioperative and surgical care pathways
- Establishing pain management protocols
- Setting patient-centered goals using SMART criteria
- Coordinating multidisciplinary care across specialties
- Ensuring regulatory compliance in treatment documentation

### Visual Enhancement

- **Vibe-Skills and OpenClaw-Medical-Skills**: Every treatment plan MUST include at least 1 AI-generated figure using the scientific-schematics skill. Generate at minimum one schematic or diagram (e.g., treatment pathway flowchart, care coordination diagram, or therapy timeline).
- **medical-research-skills**: Supports adding at least one diagram (e.g., pathway, timeline, decision algorithm) to improve usability, but it is optional.

### Document Format

- **Option 1: One-Page Treatment Plan** (PREFERRED for most cases): Single page containing all essential treatment information in scannable sections. Required sections: Header Box, Treatment Regimen, Supportive Care, Rationale, Monitoring, Evidence Level, Expected Outcome.
- **Option 2: Standard 3–4 Page Plan**: Front-page executive summary with 2–4 colored boxes (goals, interventions, decision points, timeline) plus supporting detail.
- **Option 3: Extended 5–6 Page Plan**: Only when complexity requires it (multiple comorbidities, complex monitoring, research protocols).

### Dependencies

- **Python**: 3.10+
- **TeX distribution**: TeX Live 2022+ (or MiKTeX equivalent)
- **LaTeX engines**: xelatex (recommended), pdflatex (supported)
- **Key LaTeX packages**: tcolorbox, tikz/pgf, geometry, xcolor, fontspec, fancyhdr, titlesec, enumitem, booktabs, longtable, array, colortbl, hyperref, natbib
- **Project scripts**: generate_template.py, check_completeness.py, validate_treatment_plan.py, timeline_generator.py, (optional) generate_schematic.py

### Workflow

1. Generate a LaTeX template using the bundled Python script.
2. (Optional) Generate a diagram for the plan using the scientific-schematics skill.
3. Compile to PDF using xelatex or pdflatex.
4. Run completeness and quality validation using the bundled validation scripts.
5. (Optional) Generate a timeline artifact.

### Implementation Details

- Start with the 1-page format whenever possible.
- Expand to 3–4 pages only when supporting detail is needed.
- Use 5–6 pages rarely.
- Page 1 must be a scan-friendly summary with title, patient/report info box, and 2–4 colored boxes.
- Table of contents (if used) begins on page 2; detailed sections follow.