Docs / source reviewed
JevShelf
Select skills for a task and load their instructions on demand
How this project uses Jev
- Input
- The current task, recent context, and metadata plus bounded excerpts from one configured skill collection
- Jev decides
- Uses the pinned jev-skill-router dependency to ask Jev about current-step fit and quality, returning candidates with scores, confidence information and content digests
- Code executes
- Search returns metadata; a separate load command reads selected instructions in bounded pages with optional SHA-256 continuity checks. A local lexical mode and ordinary skill files support different coding-agent hosts
Evidence and limitations
The MIT license, collection search, judgment questions, bounded loading and dependency declaration were reviewed without installation or inference. Jev mode sends task text, metadata and bounded excerpts externally. Selection does not authorize execution; the host decides whether to invoke a skill. Context savings were not measured. Private skill backup tools are a separate optional feature.
This project has not been run independently here. Author-reported results are not independently verified results.
Original sources
- https://github.com/ham-zax/jevshelf/blob/3fb8b9a35048150ffb9cbb6985e6944107d78e13/README.md ↗
- https://github.com/ham-zax/jevshelf/blob/3fb8b9a35048150ffb9cbb6985e6944107d78e13/LICENSE ↗
- https://github.com/ham-zax/jevshelf/blob/3fb8b9a35048150ffb9cbb6985e6944107d78e13/src/jevshelf/cli.py ↗
- https://github.com/ham-zax/jevshelf/blob/3fb8b9a35048150ffb9cbb6985e6944107d78e13/src/jevshelf/selection.py ↗
- https://github.com/ham-zax/jevshelf/blob/3fb8b9a35048150ffb9cbb6985e6944107d78e13/pyproject.toml ↗