CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

emerging-techniques-model-pruning

Reduce LLM size and accelerate inference using pruning techniques like Wanda and SparseGPT. Use when compressing models without retraining, achieving 50% sparsity with minimal accuracy loss, or enabling faster inference on hardware accelerators. Covers unstructured pruning, structured pruning, N:M sparsity, magnitude pruning, and one-shot methods.

Quality

68%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./cli-tool/components/skills/ai-research/emerging-techniques-model-pruning/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

38%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body delivers a genuinely executable Wanda quick start and useful comparison tables, but it is padded with basic pruning-concept explanations, its later strategies and production pipeline rely on undefined pseudocode helpers, and it inlines a large Wanda deep-dive that duplicates an existing references/wanda.md that is never linked. Main fixes: link and defer to the reference file, make the strategy/pipeline code executable, and add a validate-before-deploy checkpoint.

Suggestions

Reference references/wanda.md explicitly (e.g., a '## Wanda deep dive — See [references/wanda.md](references/wanda.md)' section) and move the duplicated Wanda criterion/algorithm details out of SKILL.md, cutting the body substantially.

Make the strategy and production-pipeline code executable: define or remove undefined helpers (train_step, fine_tune, prune_layer, prune_model, load_calibration_data, finetune_dataset), import torch.nn.functional as F for nm_prune, and pass calibration data in the format wanda_prune expects.

Add a validation checkpoint to the workflow: evaluate the pruned model (e.g., lm-eval on arc_easy/hellaswag) against a degradation threshold before saving/deploying, with a retry (lower sparsity or fine-tune) if the threshold is exceeded.

DimensionReasoningScore

Conciseness

The 485-line body includes several padded sections explaining concepts Claude already knows: 'Structured vs Unstructured' pros/cons lists, 'Sparsity Patterns' bit-mask illustrations ('[1, 0, 1, 0, 1, 1, 0, 0]... Pros: Flexible, high quality'), a baseline 'Magnitude Pruning' snippet, and pseudo-code if-statements in 'Method Selection' ('if no_retraining_budget: use_method = "wanda"'). The Wanda criterion is also explained twice (full quick-start code plus a 'Core Concepts' re-explanation). This matches anchor 2 ('several unnecessary explanations or padded sections') rather than 3, where only a small amount of trimming would be needed.

2 / 5

Actionability

The Wanda quick start is concrete and near-executable ('def wanda_prune(model, calib_data, sparsity=0.5)...'), but many later blocks are pseudocode with undefined helpers: 'train_step(model)', 'prune_layer(module, sparsity)', 'fine_tune(model, epochs=2, lr=1e-5)', 'prune_model(model, sparsity=...)', 'load_calibration_data()', and an undefined 'finetune_dataset' in the production pipeline. 'nm_prune' uses 'F.pad' without importing torch.nn.functional, the pipeline passes a HF Dataset where wanda_prune expects a list of strings, and the evaluation snippet subtracts result dicts ('(original_results - pruned_results)'). This is the anchor-3 mix of concrete guidance with pseudocode and missing key details, not 4, whose code is mostly executable with only minor gaps.

3 / 5

Workflow Clarity

The production pipeline lists a numbered sequence ('1. Load model... 2. Load calibration data... 3. Apply pruning... 4. (Optional) Fine-tune... 5. Save') and an Evaluation section exists, but pruning is a destructive, in-place batch modification of all weights ('W *= mask.float()') with no integrated validation checkpoint — the evaluation of pruned vs original is a separate section, not a step with a threshold or error-recovery loop. Per the rubric's cap for destructive/batch operations without validation, this cannot exceed 3.

3 / 5

Progressive Disclosure

The bundle contains references/wanda.md (347 lines covering Wanda's algorithm, calibration, deployment, and results) that is never mentioned in SKILL.md — it is undiscoverable — while the body inlines substantial duplicated Wanda material (full quick-start code, 'Core Concepts' pruning criteria, resources) that clearly belongs in that reference file. Scoring against the actual bundle structure, this matches anchor 2 ('content that clearly belongs in separate files is inlined; or references are buried'): the reference exists but is never signaled, and ~500 lines of content that should be split are inline. Not 1 because the body itself has clear section headers; not 3 because there is no clearly signaled reference at all and the body is monolithic.

2 / 5

Total

10

/

20

Passed

Description

83%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description that explicitly states both capability and usage triggers in third person, with concrete technique names and a well-formed 'Use when...' clause. Minor improvement is possible by adding synonyms like 'model compression' or 'sparsify' and widening the action coverage beyond reduce-size/accelerate-inference.

DimensionReasoningScore

Specificity

Concrete actions are named ('Reduce LLM size and accelerate inference using pruning techniques like Wanda and SparseGPT') and the coverage list enumerates specific technique categories ('unstructured pruning, structured pruning, N:M sparsity, magnitude pruning, and one-shot methods'). Not 5 because only two top-level actions are stated (reduce size, accelerate inference) — the rest is a topic list rather than multiple distinct actions; not 3 because several specific, concrete capabilities are explicitly listed with minor gaps.

4 / 5

Completeness

Clearly answers both 'what' ('Reduce LLM size and accelerate inference using pruning techniques...') and 'when' with an explicit, concrete trigger clause ('Use when compressing models without retraining, achieving 50% sparsity with minimal accuracy loss, or enabling faster inference on hardware accelerators'), written in third person. Matches the anchor-5 example structure exactly.

5 / 5

Trigger Term Quality

Natural user phrases are present: 'pruning', 'compressing models', 'sparsity', 'without retraining', 'faster inference on hardware accelerators', plus method names Wanda and SparseGPT. Not 5 because common synonyms such as 'model compression' or 'sparsify' are absent (N/A for file extensions); not 3 because keyword coverage is good and users would naturally say these terms when needing this skill.

4 / 5

Distinctiveness Conflict Risk

Clear niche (LLM pruning) with distinct method-name triggers (Wanda, SparseGPT, N:M sparsity). Not 5 because 'compressing models' and 'accelerate inference' overlap with closely related skills such as quantization or distillation; not 3 because the technique-specific triggers make it mostly distinguishable.

4 / 5

Total

17

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_field

'metadata' should map string keys to string values

Warning

Total

15

/

16

Passed

Repository

davila7/claude-code-templates

Path

cli-tool/components/skills/ai-research/emerging-techniques-model-pruning/SKILL.md

Reviewed

Table of Contents

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.