model-pruning-helper

Model Pruning Helper - Auto-activating skill for ML Deployment. Triggers on: model pruning helper, model pruning helper Part of the ML Deployment skill category.

Quality

Does it follow best practices?

Impact

Pending

No eval scenarios have been run

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./planned-skills/generated/08-ml-deployment/model-pruning-helper/SKILL.md

Evaluation results

100%

Reducing ResNet Model Size for Edge Deployment

Production-ready pruning with validation

Criteria

Without context

With context

Configurable pruning ratio

100%

Pruning applied to model

100%

Parameter count reported

100%

Sparsity reported

100%

Output validation step

100%

Structured step ordering

100%

pruning_report.json written

100%

Error handling present

100%

Reusable design

100%

No notebook-only patterns

100%

Uses standard ML libraries

100%

Without context: $0.4055 · 3m 32s · 23 turns · 23 in / 5,515 out tokens

With context: $0.5421 · 4m 9s · 30 turns · 988 in / 6,951 out tokens

100%

Automated Model Compression Pipeline for Monthly Retraining

MLOps pipeline with best practices

Criteria

Without context

With context

Staged pipeline structure

100%

External config parameters

100%

pipeline_config.json written

100%

Timestamped log entries

100%

Validation gate enforced

100%

Non-zero exit on failure

100%

pruned_model.pt saved on pass

100%

Failure reason logged

100%

Reusable model definition

100%

No hardcoded absolute paths

100%

Industry-standard pruning approach

100%

Without context: $0.6704 · 3m 32s · 34 turns · 35 in / 8,777 out tokens

With context: $1.3984 · 1s · 1 turns · 3 in / 20 out tokens

100%

Serving Infrastructure and Monitoring for a Pruned Recommendation Model

Model serving and monitoring post-pruning

Criteria

Without context

With context

HTTP serving endpoint

100%

Request logging with timestamps

100%

Divergence metric computed

100%

Alert threshold configurable

100%

monitoring_report.json written

100%

Alert flag set correctly

100%

Both model versions run

100%

Serving handles JSON input/output

100%

No external data dependencies

100%

Production-ready structure

100%

Standard serving framework used

100%

Without context: $0.2902 · 2m 42s · 20 turns · 20 in / 4,665 out tokens

With context: $0.4908 · 3m 16s · 29 turns · 28 in / 5,575 out tokens

Repository: jeremylongshore/claude-code-plugins-plus-skills
Commit: 0ad6004

Evaluated: 14 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Reducing ResNet Model Size for Edge Deployment Automated Model Compression Pipeline for Monthly Retraining Serving Infrastructure and Monitoring for a Pruned Recommendation Model

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.