A token efficient, local LLM friendly agentic coding harness: a complete agent runtime (agent loop, tool pipeline, sandboxing, session persistence) with four front-ends — a Blazor web UI, a headless CLI, and JSON-RPC / ACP stdio servers for editors and automation.
Agent tooling — page 19
Frameworks, routers and utilities that use Jev to choose tools or branches inside a larger agent system.
1,206 active projects in this collection. All collections →
微信聊天内容记录搜索mcp
Workload-aware Pareto frontiers and dynamic model selection for agentic systems
Community plugins, developer tooling, and shared utilities for extending DeepSeek Harness — built as an Nx + pnpm monorepo.
2
A personal Claude Code harness: doctrine, mechanical gates, and curated skills. Holds no personal data by design.
Local-first chat agent that runs comfortably with small-context local LLMs (from 2B) without losing the original goal of a long conversation. Multi-window context with rollovers, SKILL.state runtime, memory carry-over, three-tier auto web search, and a stalled-session detector. Next.js + FastAPI + SQLite + LanceDB, packaged for Docker Compose.
2
Agentic RAG pipeline for biomedical evidence discovery: query planning and multi-round search across twelve public databases plus citation-graph expansion, ranking by relevance and study design, and citation-grounded synthesis with disagreements reported explicitly.
Deterministic, query-aware context compaction — harness-agnostic, local, reversible, fail-open, multilingual, and zero-LLM in the extractive path.
AI Harness - JIT policy routing, MCP/Skill capability acquisition, shared machine runtime registry
Quota-aware subagent routing for Claude Code. Reads live rate limits across codex/grok/kimi, dispatches to the CLI with headroom, isolates in a git worktree, and verifies the diff before reporting.
No project description available.
Agent Skills for Claude Code — spec-driven OpenSpec + TLA+ sub-agent workflow
Agent skill biên tập tiếng Việt tự nhiên hơn, giữ nguyên ý, dữ kiện và giọng người viết.
SkyN3t 2.0 — autonomous multi-agent app factory
Open-source MCP and HTTPS screening for prompts, skills, and tool descriptions using TypeSafe AI Jev, deterministic checks, and configurable reasoning.
Long-term memory for AI agents on Jev. Keep raw records, judge them with a fast System 1 model and answer from under 4k tokens of context.
Route smarter. Code in parallel. Resume what breaks. A Jev-powered multi-agent factory for Codex.
Review independente sobre a adequação do Jev a produtos e desenvolvimento multiagente
Route each coding-agent turn to the cheapest model that can finish it. Switchyard-style execution signals, TypeSafe's jev as the judge, honest cost accounting. Claude Code, Codex, OpenCode, Pi.
MCP server and skill that let an agent drive an Android phone naturally over adb
VS Code Codex Plugin for Jev-gated tool output integration
Tool search for LLM agents: BM25 vs embeddings vs rerankers vs Jev on 525 real MCP tools, plus an experimental Jev search engine
Temporal evidence memory. Reconstruct historical context and trace decisions affected by corrections.
JEV and local LAYA for MCP. Typed decisions, bounded inference and durable decision records.