pddl-skills

Load PDDL domains and problems, generate plans with classical planners, and validate and save them.

Updated Jan 15, 2026
One-click install
npx skills add https://github.com/KaiserWhoLearns/skillsbench --skill pddl-skills
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pddl-skills
Source: https://github.com/KaiserWhoLearns/skillsbench/tree/main/tasks/pddl-robot-planning/environment/skills/pddl-skills
Command: npx skills add https://github.com/KaiserWhoLearns/skillsbench --skill pddl-skills

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires unified_planning.

What problem does it solve?

Automated planning utilities for loading PDDL domains and problems, generating plans using classical planners, validating plans, and saving plan outputs.

Core Features & Use Cases

  • Load Domain and Problem: parse domain and problem files into a unified planning problem.
  • Plan Generation: generate a plan using a classical planner.
  • Plan Validation: verify plan correctness against the problem.
  • Save Plan: persist the plan to disk for reproducibility.
  • Use Case: benchmark different planners for PDDL tasks and reproduce results.

Quick Start

Load a PDDL domain and problem with the PDDL reader, generate a plan with a classical planner, validate the plan, and save the resulting plan to disk.

Frequently Asked Questions about pddl-skills

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate and validate a PDDL plan using a classical planner?

To generate and validate a PDDL plan, load domain and problem files into a unified planning problem, generate a plan with a classical planner, verify correctness, and save the resulting plan to disk.

What is the unified_planning library used for in automated planning workflows?

The unified_planning library provides a PDDLReader to parse domain and problem files and a OneshotPlanner interface to generate plans, enabling end-to-end automated planning pipelines for reproducible experiments.

Do I need the unified_planning library to parse PDDL domain and problem files?

Yes, this Skill requires the unified_planning library and its PDDLReader component to load PDDL domains and problems, generate plans, and validate the resulting automated planning outputs.

Can I benchmark different classical planners for PDDL task planning?

You can benchmark different planners for PDDL tasks by loading domains and problems, generating plans with classical planners, validating correctness, and saving outputs to reproduce results.

How do I save a validated PDDL plan to disk for reproducible experiments?

After generating and validating a PDDL plan, the workflow persists the serializable plan to disk for verification and storage, ensuring reproducible planning experiments and benchmarking results.