algo-nlp-summarization
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill describes a standard NLP pipeline for document summarization using extractive and abstractive methods. The provided code snippets are well-documented and utilize established libraries like transformers and spacy.
- [EXTERNAL_DOWNLOADS]: The skill references machine learning models hosted on Hugging Face and official SpaCy models. These are well-known services and trusted sources for NLP resources.
- [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted document input for summarization. Ingestion points: Input text processed in SKILL.md and references/factual-consistency.md. Boundary markers: Hallucination gates and validation steps are described in the algorithm phases. Capability inventory: Subprocess calls or network writes are absent; logic is limited to text analysis and model inference. Sanitization: Token truncation and factual consistency scoring (NLI/QA) are integrated into the pipeline.
Audit Metadata