pymoo
by K-Dense-AIpymoo is a Python skill for single- and multi-objective optimization, Pareto fronts, constrained problems, and benchmark tests. Use this pymoo guide to choose algorithms like NSGA-II, NSGA-III, and MOEA/D, follow the install and usage workflow, and apply pymoo for Data Analysis when multiple metrics must be balanced.
This skill scores 78/100, which means it is a solid listing candidate for directory users: it clearly targets a real optimization workflow, gives enough structure for an agent to recognize when to use it, and provides substantial operational guidance, though it would still benefit from more adoption aids like install instructions and supporting files.
- Clear, specific trigger conditions for optimization tasks, including Pareto fronts, constraint handling, and benchmark problems.
- Substantial workflow content with a large skill body, many headings, and code examples, which should reduce guesswork for agents.
- Well-scoped to a recognizable Python optimization framework with concrete algorithms and problem types named in the description.
- No install command and no support files (scripts, references, resources, or rules), so users must infer some adoption details from the SKILL.md alone.
- The repository appears to be documentation-only for this skill, so its practical leverage depends on whether the agent already has the underlying Python environment and pymoo library available.
Overview of pymoo skill
pymoo is a Python skill for solving single- and multi-objective optimization problems with a strong focus on Pareto trade-offs, evolutionary algorithms, and constrained design problems. Use the pymoo skill when you need more than a generic optimizer: it helps you choose and configure algorithms like NSGA-II, NSGA-III, and MOEA/D, then evaluate results in a way that matches real engineering or data analysis decisions.
This is best for readers who already have an objective function, constraints, and a need to compare competing outcomes rather than optimize one number. If your job is to find feasible, high-quality trade-off solutions, the pymoo skill is a good fit.
What pymoo is for
pymoo is built for multi-objective optimization workflows: selecting algorithms, defining problem structure, running optimization, and interpreting Pareto fronts. It also supports benchmark problems such as ZDT and DTLZ, so it is useful for both applied work and method comparison.
Who should use this skill
Use the pymoo skill if you are:
- modeling engineering design problems with competing objectives
- benchmarking optimization methods
- exploring constrained optimization in Python
- doing pymoo for Data Analysis where multiple metrics must be balanced
- comparing solution sets instead of a single best answer
Why pymoo stands out
The main value is the unified optimization workflow: one framework, multiple algorithm families, and consistent result handling. That reduces setup friction when you move from a small test problem to a more realistic constrained problem. It is especially helpful when you need to inspect trade-offs instead of forcing everything into one score.
How to Use pymoo skill
Install pymoo skill
Install the skill in your agent environment with the repository path:
npx skills add K-Dense-AI/claude-scientific-skills --skill pymoo
After installation, read scientific-skills/pymoo/SKILL.md first. This is the most important file because it defines the intended workflow, core concepts, and constraints for using pymoo correctly.
Start from the right input
The pymoo skill works best when your prompt includes:
- decision variables and their bounds
- number of objectives
- constraints, if any
- whether variables are continuous, discrete, binary, or mixed
- your target output: best solution, Pareto set, comparison, or benchmark run
A weak prompt says: “Use pymoo to optimize my model.”
A stronger prompt says: “Use pymoo to solve a constrained two-objective problem with 8 continuous variables, minimize cost and error, and return the Pareto front plus a recommended knee-point solution.”
Practical workflow
A good pymoo usage flow is:
- define the problem clearly
- choose an algorithm that matches the objective count and variable type
- run
minimize()with explicit termination settings - inspect feasibility, convergence, and Pareto quality
- compare candidate solutions with your real business or engineering criteria
This matters because the quality of the setup usually matters more than the algorithm name. Many poor results come from underspecified objectives or missing constraints, not from pymoo itself.
Read these files first
For this skill, start with:
scientific-skills/pymoo/SKILL.md
Then scan the sections covering:
- when to use the skill
- core concepts
- unified interface
- constraints and practical workflow examples
Because there are no extra support files in this repo, the main source of truth is the skill document itself.
pymoo skill FAQ
Is pymoo only for multi-objective problems?
No. pymoo supports both single- and multi-objective optimization, but its real strength is handling trade-offs cleanly. If you only need basic scalar minimization, a simpler tool may be enough.
Is this better than a generic prompt for optimization?
Usually yes. A generic prompt may produce a plausible algorithm choice, but the pymoo skill gives you a consistent workflow, explicit terminology, and better guidance for constraints, Pareto fronts, and benchmark problems. That reduces guesswork when you need reproducible pymoo usage.
Is pymoo beginner-friendly?
It is beginner-friendly if you already understand the optimization problem you want to solve. It is less beginner-friendly if you do not know your variables, constraints, or success metric yet. The skill helps most when the problem is defined but the implementation details are still unclear.
When should I not use pymoo?
Do not use pymoo if your task is plain data wrangling, visualization, or standard statistical analysis with no optimization step. For pymoo for Data Analysis, it only makes sense when you are optimizing a model, feature set, threshold, or policy across multiple criteria.
How to Improve pymoo skill
Give the model a fully specified problem
The biggest improvement comes from clearer problem framing. Include objective formulas, variable ranges, constraints, and what “good” looks like in practice. If you want a Pareto front, say so; if you want one deployable solution, say how to choose it.
Match the algorithm to the problem
Tell the prompt what kind of search you need. For example, mention whether you want NSGA-II for a common multi-objective setup, or whether your problem has discrete or mixed variables. That helps the pymoo skill avoid generic recommendations that do not fit your variable space.
Ask for decision-ready output
If you only ask for code, you may get a working script but not a usable outcome. Better prompts request the optimization setup, termination criteria, result interpretation, and a short note on how to validate the solution set. This is especially important for pymoo install and usage scenarios where you are integrating the skill into an existing Python workflow.
Iterate with constraints and trade-offs
If the first result is too broad, refine it by adding:
- tighter bounds
- explicit feasibility rules
- preferred trade-off direction
- performance thresholds for acceptable solutions
That kind of iteration improves pymoo guide quality because the skill can focus on solution ranking instead of guessing your selection criteria.