sentence-transformers

Framework for state-of-the-art sentence, text, and image embeddings. Provides 5000+ pre-trained models for semantic similarity, clustering, and retrieval. Supports multilingual, domain-specific, and multimodal models. Use for generating embeddings for RAG, semantic search, or similarity tasks. Best for production embedding generation.

Quality

82%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

65%

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

A code-rich, highly actionable reference for sentence-transformers, weakened by duplicated model content across three inline sections plus an unreferenced bundle file, and by missing validation checkpoints in the fine-tuning workflow. Trimming the redundant model tables, linking references/models.md, and adding a save/validate step would lift the lower dimensions.

Suggestions

Consolidate the 'Popular models', 'Model selection guide', and 'Performance' tables into a single source and link out to references/models.md instead of duplicating model specs inline; remove the GitHub-stars/marketing metrics block.

Reference references/models.md explicitly from the body (e.g., 'See [models.md](references/models.md) for the full model catalog') so the bundle file is actually discoverable.

Add a validation/checkpoint step to the fine-tuning workflow (e.g., reload the saved model and verify embeddings shape, or evaluate on a held-out pair) to close the workflow-clarity gap.

Dimension	Reasoning	Score
Conciseness	Code blocks are lean and action-forward, but model information is duplicated across the 'Popular models', 'Model selection guide', and 'Performance' sections (and again in references/models.md), and the '15,700+ GitHub stars' / '5000+ pre-trained models' metrics are marketing fluff Claude cannot act on; mostly efficient but could be tightened rather than fully lean.	2 / 3
Actionability	Provides abundant executable, copy-paste Python with specific model identifiers across installation, encoding, semantic search, batch processing, fine-tuning, and LangChain/LlamaIndex integration, matching the 'fully executable...copy-paste ready' anchor.	3 / 3
Workflow Clarity	Sequences are present (install → load → encode; prepare data → dataloader → loss → fit → save), but multi-step operations like fine-tuning lack explicit validation checkpoints or feedback loops, fitting 'steps listed but validation gaps' rather than the validated level-3 anchor.	2 / 3
Progressive Disclosure	The body is well-sectioned and a bundle file references/models.md exists, but it is never linked or signaled from the body and its model-guide content is duplicated inline — 'references present but not clearly signaled; content that should be separate is inline' — so it does not reach the well-signaled one-level-deep level-3 anchor.	2 / 3
	Total	9 / 12 Passed

Description

100%

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, third-person description that states concrete capabilities and pairs them with an explicit 'Use for...' trigger clause covering RAG, semantic search, and similarity. It would benefit from trimming the 'state-of-the-art' buzzword and 'Best for production embedding generation' over-claim, but these do not materially weaken its specificity or triggers.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions and capabilities — 'sentence, text, and image embeddings', 'semantic similarity, clustering, and retrieval', 'generating embeddings for RAG, semantic search' — matching the 'lists multiple specific concrete actions' anchor rather than the partial level 2.	3 / 3
Completeness	Explicitly answers both 'what' ('Framework for...embeddings...Provides 5000+ pre-trained models for...') and 'when' via the explicit trigger clause 'Use for generating embeddings for RAG, semantic search, or similarity tasks', so it is not capped at 2.	3 / 3
Trigger Term Quality	Contains natural terms a user would actually say when needing this skill — 'embeddings', 'RAG', 'semantic search', 'similarity', 'clustering' — giving good coverage; the 'state-of-the-art' buzzword is incidental framing rather than a missing keyword.	3 / 3
Distinctiveness Conflict Risk	Occupies a clear embedding-generation niche with distinct triggers (RAG / semantic search / similarity) in third-person voice, making it unlikely to fire for unrelated skills; not merely 'somewhat specific' as at level 2.	3 / 3
	Total	12 / 12 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: Orchestra-Research/AI-Research-SKILLs
Path: 15-rag/sentence-transformers/SKILL.md
Commit: 773a529

Reviewed: about 5 hours ago

Table of Contents

Discovery Implementation Validation

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.