umap-learn
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes external data arrays for dimensionality reduction and visualization.\n
- Ingestion points: Data is passed to the
fit(),fit_transform(), andtransform()methods acrossSKILL.mdandreferences/api_reference.md.\n - Boundary markers: Not present in provided code snippets.\n
- Capability inventory: Includes data transformation, plotting, and model persistence.\n
- Sanitization: Utilizes
ensure_all_finitevalidation for input data as described in the API reference.\n- [DYNAMIC_EXECUTION]: The skill leverages JIT compilation and neural network execution for its core functionality.\n - Evidence: Employs Numba's
@njitdecorator for custom distance metrics and TensorFlow/Keras models for Parametric UMAP as documented inreferences/advanced-features.md.\n- [EXTERNAL_DOWNLOADS]: Fetches standard, well-known data science libraries from official package registries.\n - Evidence: Mentions installation of
umap-learnandhdbscanusing theuvpackage manager inSKILL.md.
Audit Metadata