User repository on GitHub
Jev Skill MCP Router — samler skill-kilder, indekserer dem, lader Jev vælge de bedste til et projekt
- Forks
- 0
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Agents & MCP
Jev Skill Router — king4s
A Rust MCP server that recommends third-party agent skills: it retrieves deduplicated metadata with SQLite FTS5/BM25, sends one comparable typed Score per candidate in batches of at most 128 questions, and never installs or executes a skill.
- Stars
- 3
- Language
- Python
- Reading time
- 2 min
- Published by
- Community
- Author
- King4s
- Last verified
- Oct 2026
This is one of several Jev listings for the same project. The canonical entry is jev-skill-router (himomohi).
Works forJev for Ranking
Used inJev for Software & DevTools
- Capabilities
- Open sourceRuns locallySelf-hosted
Highlights
- Native Rust MCP server that ranks third-party agent skills for a project and never installs or executes them.
- Pipeline: GitHub metadata, SQLite FTS5/BM25 dedup, typed decision provider, ranked recommendations.
- It sends one comparable typed Score per candidate in batches of at most 128 questions to one configured provider.
- The Rust service opens SQLite read-only and calls the selected provider directly, with no Python wrapper or shell subprocess.
- The Python indexer with PyYAML and the evaluation scripts are offline tooling; GitHub metadata lists no license and a read-only safety badge.
Watch out
README and GitHub metadata reviewed 2026-10-17; the excerpt truncates mid-sentence in the architecture note, so the full provider options, install steps and verification workflow are not evidenced. Created 2026-10-04, last pushed 2026-10-13.
Questions about this entry
Short answers drawn from the linked project or page.
A native Rust MCP server that retrieves skill metadata from GitHub, indexes it with SQLite FTS5/BM25, and ranks candidates with a selected decision provider, publishing the result as ranked skill recommendations for a project.
It retrieves deduplicated metadata, then sends one comparable typed Score per candidate, in batches of at most 128 questions, to a single configured provider, and turns the returned scores into ranked recommendations.
No. The README says it never installs or executes skills, and the repository carries a read-only safety badge; the Rust service also opens its SQLite database read-only.
The repository description says Jev chooses the best skills for a project, and the README describes a selected decision provider receiving typed Score questions. The digest does not show which providers can be configured.
The running service is Rust and is exposed as an MCP server. The Python indexer, which uses PyYAML, and the evaluation scripts are offline tooling; the Rust service does not wrap Python or spawn a subprocess.
The README advertises read-only behavior: it does not install or execute the skills it ranks, and it opens SQLite read-only while calling the configured provider directly, which limits its side effects to the provider call.
GitHub metadata shows 3 stars, created 2026-10-04 and last pushed 2026-10-13, with a verification workflow badge in the README. No license is listed in the digest metadata.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
X: Jev in front of the HeyGen MCP
X: Playwright CLI + Jev vs Playwright MCP
X: A local MCP server that cuts computer use from 23s to 0.8s. Cartwright built Cortex, a local MCP server where…
X: A side panel that drives any site's WebMCP tools
More Jev resources like this
Local-first MCP memory server for IDEs: tree-sitter indexing into a local ChromaDB store, plus an optional Jev layer that screens retrieved passages for relevance, evidence, contradiction and prompt injection.
Jev Agents & MCP#community#github#python
Community
KikeVen/zerikai_memoryPython
KikeVen/zerikai_memoryPythonLocal code search for coding agents: ripgrep and BM25 shortlist function-level chunks, Jev Noul relevance questions run across parallel requests, and accepted excerpts return to Codex, Claude Code, OpenCode or Pi through MCP or the `oko ask` command.
Jev Agents & MCP#community#mcp#code-search
Community
bartlomein/okoRust
bartlomein/okoRustA self-hosted MCP server that uses TypeSafe JEV as a semantic evidence-selection layer for coding-agent web research: it filters search results before fetching and parsed content blocks afterward, returning compact evidence blocks with provenance.
Jev Agents & MCP#community#github#mcp
Community
merth1470/jev-research-mcpPython
merth1470/jev-research-mcpPythonA local System One decision layer for AI coding agents, shipped as an MCP server and Agent Skill, that offloads high-volume classification, checks, scores and ranking from frontier LLMs to sub-50ms local models with calibrated output.
Jev Agents & MCP#community#github#mcp
Community
aktonay/snapdecPython
aktonay/snapdecPythonA lightweight self-hosted System 1 decision daemon and MCP server for AI agents and multi-agent swarms: it evaluates state text against typed choice, score and noul questions in one forward pass, with Laya, Ollama and OpenAI-compatible backends.
Jev Agents & MCP#community#github#mcp
Community
axiomantic/local-systemonePython
axiomantic/local-systemonePythonAn MCP server and CLI web crawler for AI agents in which a small decision model picks every link and passage, so the agent's LLM reads only what answers its question. On 50 WebWalkerQA tasks the README claims 40% correct answers at about $0.025 per task.
Jev Agents & MCP#community#github#python
Community
rayenmlayeh/jevcrawlPython
rayenmlayeh/jevcrawlPythonFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.