CtrlK
BlogDocsLog inGet started
Tessl Logo

building-classification-models

Build and evaluate classification models for supervised learning tasks with labeled data. Use when requesting "build a classifier", "create classification model", or "train classifier". Trigger with relevant phrases based on skill purpose.

40

Quality

41%

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 ./plugins/ai-ml/classification-model-builder/skills/building-classification-models/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

15%

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

The skill body is a high-level narrative about classification modeling with no executable guidance, no real workflow with validation checkpoints, and no links to its own bundled references, scripts, or assets. It reads as a description of capability rather than actionable instructions Claude can follow.

Suggestions

Add concrete, executable guidance — e.g., a minimal Python/scikit-learn training snippet or the exact command to run model_builder.py with a config — instead of narrating 'the skill utilizes the plugin to generate code'.

Turn 'How It Works' into a real sequenced workflow with a validation checkpoint (validate the dataset with data_validator.py, train, then evaluate and report) so model-training failures are caught early.

Wire the bundle into the body: link to scripts/model_builder.py, scripts/data_validator.py, and assets/model_config_template.json from the relevant sections so progressive disclosure is actually used rather than ignored.

DimensionReasoningScore

Conciseness

The body is mostly efficient prose, but sections like Overview, How It Works, Integration, Instructions, and Output restate what a classifier is and what the skill does in general terms ('automates the process of model selection, training, and evaluation') rather than giving only what Claude would not already know.

2 / 3

Actionability

The body describes and narrates rather than instructs — there is no executable code, no concrete commands, no copy-paste example, and even references to the bundled model_config_template.json and scripts are not actually invoked or shown in the body.

1 / 3

Workflow Clarity

Although 'How It Works' lists steps, they are abstract ('Context Analysis', 'Model Generation', 'Evaluation and Reporting') with no concrete sequence, commands, or validation checkpoints; for batch/model-training operations that can silently produce bad models, the missing validate-and-verify feedback loop caps this at 1.

1 / 3

Progressive Disclosure

The body is a monolithic wall of generic text and never signals the bundled references, scripts, or assets — the README files and model_config_template.json exist but are not linked to or navigated from the body, so there is no real disclosure structure.

1 / 3

Total

5

/

12

Passed

Description

67%

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

The description is well-structured with an explicit 'Use when' trigger clause and a clear, focused domain. Its main weaknesses are a thin action list, near-identical trigger phrases that miss natural variations, and a generic boilerplate trigger line that dilutes distinctiveness.

Suggestions

Expand the capability list beyond 'build and evaluate' to concrete actions like 'train, tune hyperparameters, and report accuracy/precision/recall/F1' to reach specificity level 3.

Replace the generic 'Trigger with relevant phrases based on skill purpose' with concrete natural-language variations users would say, e.g. 'classify data', 'predict categories', or 'train a spam detector'.

Vary the trigger phrases so they aren't three paraphrases of the same idea; include outcome-oriented phrasing to reduce overlap with general ML skills.

DimensionReasoningScore

Specificity

Names a domain (classification models) and a couple of actions (build, evaluate), but the action list is thin — it stops short of the multiple concrete actions (e.g., train, tune hyperparameters, generate metrics) that a level 3 example like 'Extract text and tables... fill forms, merge documents' shows.

2 / 3

Completeness

It answers both 'what' (build and evaluate classification models) and 'when' with an explicit 'Use when requesting...' clause plus trigger phrases, matching the level-3 example structure.

3 / 3

Trigger Term Quality

It includes relevant phrases ('build a classifier', 'create classification model', 'train classifier'), but these are near-duplicates of one another and miss common natural variations like 'classify data', 'predict categories', or 'spam/label detection'.

2 / 3

Distinctiveness Conflict Risk

The domain is fairly specific to classification (supervised learning with labeled data), but the boilerplate line 'Trigger with relevant phrases based on skill purpose' is generic and the 'build a classifier' triggers could still overlap with broader ML/data-science skills.

2 / 3

Total

9

/

12

Passed

Validation

87%

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

Validation14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

allowed_tools_field

'allowed-tools' contains unusual tool name(s)

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

14

/

16

Passed

Repository
jeremylongshore/claude-code-plugins-plus-skills
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.