What it is
Spend fewer tokens in long agent sessions without changing the workflow.
What it does: Compacts stale tool output, routes subagents, and trims skill ads via Jev.
Why it belongs in the showcase: Integrated hosted-Jev optimization pack with public docs and stars.
Caveats: Configure carefully; fail-open.
Evidence
Public MIT repo, created 2026-10-08, 50 stars as of 2026-10-30. README documents jev-compacter, jev-router, and jev-skills as independent tools, Claude Code install path, Codex notes, architecture docs, and demo assets. Broader productized pack than single-purpose jev-pruner / hermes-jev-routing.
Caveats
- Measured cost reports remain Claude-only per README
- Router/skills defaults may be off; read /config
- Overlaps conceptually with smaller single-purpose repos already listed
Unofficial case page. Stars / list position observed for heat; do not treat heat as a quality SLA. Review the upstream README before you depend on it.