agentleFS
LocloveLoclove/cumcm-skills/skills/cumcm-model-planner/SKILL.md

cumcm-model-planner

Skill4 starsChanged 3 months agoMITView on GitHub

Select and compare algorithms for CUMCM mathematical-modeling problems, including optimization, forecasting, evaluation, simulation, graph/network, image processing, differential equations, machine learning, and hybrid physics-data models. Use when the user asks which model to use, how to design the method, how to validate results, or how to build a robust modeling route.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read the problem packet from cumcm-problem-reader or build a short packet if missing.

    line 12
  2. Very specific:

    Use scripts/choose_algorithm.py for a deterministic first pass when the text is long.

    line 14
  3. Distinctive:

    Do not select a black-box model without an interpretability and validation plan.

    line 48
  4. Distinctive:

    Do not use deep learning when data scale, time, or explanation needs make it unjustified.

    line 49
  5. Distinctive:

    Do not optimize an objective that is not tied to the problem's deliverables.

    line 50

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

CUMCM Model Planner

This skill chooses algorithms by problem structure, not by trendiness. It must produce a route that can be implemented, validated, explained, and written in a contest paper.

Workflow

  1. Read the problem packet from cumcm-problem-reader or build a short packet if missing.
  2. Load references/algorithm-routing.md and score candidate model families.
  3. Use scripts/choose_algorithm.py for a deterministic first pass when the text is long.
  4. Build 2-3 routes: conservative baseline, contest-grade primary model, and optional extension model.
  5. For each route, define assumptions, variables, objective/constraints, data needs, validation, sensitivity analysis, and paper figures.
  6. Reject models that cannot be explained, reproduced, or validated under contest time.
  7. Hand off to cumcm-data-loop, cumcm-figure, and cumcm-paper-writing.

Output Contract

Algorithm shortlist
| Route | Model family | Why it fits | Data needs | Validation | Risk | Implementation cost |

Recommended route
- Baseline:
- Main model:
- Enhancement:
- Why not alternatives:

Mathematical formulation
- Decision variables / states:
- Objective:
- Constraints:
- Key equations:
- Parameters:

Validation plan
- Train/test or scenario split:
- Metrics:
- Robustness and sensitivity:
- Ablations:

Red Lines

  • Do not select a black-box model without an interpretability and validation plan.
  • Do not use deep learning when data scale, time, or explanation needs make it unjustified.
  • Do not optimize an objective that is not tied to the problem's deliverables.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.