Find the tool class your 
job actually needs.
Describe the work. A local MiniLM model compares it with twelve specialist browser capabilities, then exposes the scores, evidence, and limits behind the match.
Ready for a concrete job. The pinned quantized model is bundled with this site and runs in the browser.
No analysis yet
Run the bundled sample or describe your own job to compute a ranked field.
Similarity is not a product endorsement, benchmark, security audit, or price check.
The ranked field
Open any band to inspect what was compared. Scores measure description proximity only; every production claim still needs verification.
Run an analysis to reorganize the twelve capability bands.
Why specialists surface
A general chatbot can explain many tasks. These libraries are candidates when the browser must perform the transformation itself.
The model maps meaning, not merit. A high cosine score says “these descriptions point in a similar direction.” It does not say “ship this library.” Inspect the capability, limits, maintenance, license, performance, and security before adoption.