OpenSkillIndex← back to search
claude-code

LLMs-Universal-Life-Science-and-Clinical-Skills- Plan_and_Solve

<!--

install
source · Clone the upstream repo
git clone https://github.com/mdbabumiamssm/LLMs-Universal-Life-Science-and-Clinical-Skills-
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/mdbabumiamssm/LLMs-Universal-Life-Science-and-Clinical-Skills- "$T" && mkdir -p ~/.claude/skills && cp -r "$T/Skills/Agentic_AI/Agent_Architectures/Plan_and_Solve" ~/.claude/skills/mdbabumiamssm-llms-universal-life-science-and-clinical-skills-plan-and-solve && rm -rf "$T"
manifest: Skills/Agentic_AI/Agent_Architectures/Plan_and_Solve/SKILL.md
source content
<!-- # COPYRIGHT NOTICE # This file is part of the "Universal Biomedical Skills" project. # Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu> # All Rights Reserved. # # This code is proprietary and confidential. # Unauthorized copying of this file, via any medium is strictly prohibited. # # Provenance: Authenticated by MD BABU MIA -->

name: 'plan-and-solve-agent' description: 'Breaks down complex queries into a step-by-step plan before execution, improving performance on multi-hop reasoning tasks.' measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes. allowed-tools:

  • read_file
  • run_shell_command

Plan-and-Solve Agent

The Plan-and-Solve Agent separates high-level planning from low-level execution. It is ideal for complex scientific inquiries that require multiple distinct steps (e.g., "Find targets for disease X, then design drugs, then check safety").

When to Use This Skill

  • When a user query is too complex for a single "ReAct" loop.
  • When you need to visualize the reasoning process before committing to execution.
  • To orchestrate multiple specialized sub-agents.

Core Capabilities

  1. Decomposition: Splits a goal into linear or parallel sub-tasks.
  2. Execution: runs each step sequentially (mocked in this version).
  3. Reporting: Summarizes the outputs of all steps.

Workflow

  1. Input: A complex natural language query.
  2. Plan: The agent generates a list of 
    PlanNode
    objects.
  3. Execute: The agent iterates through nodes, executing them (simulation).

Example Usage

User: "Investigate the impact of variant X on drug response."

Agent Action:

python3 Skills/Agentic_AI/Agent_Architectures/Plan_and_Solve/plan_and_solve.py \
    --query "Investigate the impact of variant X on drug response."



<!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->