Menu
This skill should be used when the user asks to tune parameters, run hyperparameter search, optimize algorithm settings, or find the best configuration for a computational math method.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
| name | auto-tuning-skill |
| description | This skill should be used when the user asks to tune parameters, run hyperparameter search, optimize algorithm settings, or find the best configuration for a computational math method. |
Use this Skill after a computational math reproduction has succeeded or partially succeeded and the human has approved tuning. Codex proposes the tuning plan from the reproduction evidence, explains the parameter space and budget, and waits for approval before running experiments.
This Skill is not a first entrypoint and is not launched through a user-facing CLI pipeline. For end-to-end work, computational_math_reproduction_workflow_skill selects this Skill when tuning becomes appropriate.
Tuning must not start until:
tuning/tuning_plan.md;Budget increases, expanded parameter spaces, dependency changes, and source edits require another approval.
Put tuning artifacts under the run directory:
tuning/
├── tuning_plan.md
├── tuning_results.csv
├── best_parameters.json
├── tuning.log
├── tuning_figures/
└── TUNING_SUMMARY.md
skills/auto_tuning_skill/scripts/experiment_runner.py is an optional helper for repeatable grid search or random search after approval. It is a tool Codex may call, not the workflow driver.
Typical helper use:
python -m skills.auto_tuning_skill.scripts.experiment_runner \
--source /path/to/repo \
--param-space /path/to/parameter_space.yaml \
--budget 20 \
--out outputs/<run_id>/tuning \
--method random_search \
--require-approval tuning_plan
The helper writes tuning_results.csv and best_parameters.json. Codex should also keep a concise tuning.log, generate figures under tuning/tuning_figures/ when metrics are available, and write TUNING_SUMMARY.md.
tuning/tuning_plan.md with parameter space, search method, budget, objective metric, constraints, stop conditions, and risks.approve, revise, reject, or skip.experiment_runner.py --require-approval tuning_plan.tuning_results.csv, best_parameters.json, tuning.log, and any figures.tuning/TUNING_SUMMARY.md with best parameters, metric evidence, failed trials, and whether the result changes the reproduction conclusion.Route AI4Math computational mathematics tasks to the appropriate bundled skill package.
Use when reproducing, deploying, diagnosing, tuning, visualizing, or reporting computational math research code with a coding agent
This skill should be used when the user asks to search for external algorithms, find related implementations, discover alternative approaches, or look up papers and code for computational math optimization methods.
Use when classifying a computational math repository or task into a broad domain before choosing specialist Skills, references, runtimes, and validation evidence.
Use when a user starts an end-to-end computational math research-code reproduction, deployment, tuning, visualization, or reporting workflow with a coding agent.
This skill should be used when the user asks to detect what algorithm family a piece of code uses, identify optimization methods, or classify computational math approaches such as ADMM, PPA, proximal gradient, or primal-dual methods.