jev-skill-suggester: a lightweight pre-execution skill router stops agents picking wrong tools
udmrzn · x · 2026-09-21
Open-source jev-skill-suggester adds a lightweight, non-invasive skill-routing layer for agents overloaded with similar skills: before execution, it recommends which SKILL.md the agent should read first.
- Two-stage filtering: screens skill descriptions, then verifies body snippets; only recommends at ≥0.80 fit with sufficient confidence
- Safe isolation: standalone Python CLI (stdlib only), doesn't modify agent config or execute skills — the host agent keeps final say
- Tested and cheap: 15/16 preset tasks matched expectations, 5.05s median per task, under $0.01 total API cost, with a 24h local response cache
- Supports an offline local mode (keyword candidates only) and a Jev semantic mode
More from coding & agent
- Dev Launches Made With Jev, a Free Directory Cataloging Demos, Tools and Skills for the New Model — Sea_Supermarket_5891 · 2026-09-21
- User's Persistent Grok Agent Buys Raw Milk While They Sleep — RachelVT42 · 2026-09-21
- Which Model Actually Understands Reverse Engineering? Dev Seeks MCP Workflow for Ghidra and IDA Pro — obese_coder · 2026-09-21
- Graphify + Obsidian + Claude Code: turn file piles into a browsable knowledge graph for agents — alex_verem · 2026-09-21
- One-time Gemini video context extraction powers faster, more accurate search — eptwts · 2026-09-21
- Claude Code bug leaves stale worktreePath metadata after worktree directories are removed — Aadiboy21 · 2026-09-21