spacy
spaCy NLP library with pipelines. Use for text processing.
Install
npx skillscat add g1joshi/agent-skills/spacy Install via the SkillsCat registry.
We need to produce a 2-3 sentence plain-text summary, objective, factual, no marketing language, no superlatives, no calls to action. Must be at most 60 words. Must not wrap in quotes. Must be only the summary text. No markdown, no bullet points, no headings. So just plain text sentences. We need to explain what the skill does, what problem it solves, when to use it. We can produce something like: "The skill provides a spaCy NLP library that enables text processing tasks such as tokenization, part‑of‑speech tagging, dependency parsing, and named entity recognition via pipeline components.
spaCy
spaCy is "Industrial Strength" NLP. Unlike NLTK (academic), spaCy focuses on providing the best single algorithm for a task. v3.8 supports Python 3.13.
When to Use
- NER (Named Entity Recognition): Extracting person names, dates, orgs.
- Parsing: Dependency parsing to understand sentence structure.
- Speed: Cython-optimized pipelines.
Core Concepts
Pipeline
Tokenizer -> Tagger -> Parser -> NER.
Doc / Token / Span
The core data structures. Efficient memory usage.
Prodigy
The annotation tool (paid) from the same creators, tightly integrated.
Best Practices (2025)
Do:
- Use Transformer pipelines:
en_core_web_trf(Roberta-based) for high accuracy. - Use
nlp.pipe(): For batch processing huge texts.
Don't:
- Don't use for GenAI: spaCy is for structure extraction, not text generation (LLMs).
References
Categories
Install
npx skillscat add g1joshi/agent-skills/spacy Install via the SkillsCat registry.