Verification-gated skill routing and self-improvement harness for Hermes-style agent skills
-
Updated
Jul 15, 2026 - Python
Verification-gated skill routing and self-improvement harness for Hermes-style agent skills
The skill OS for Codex, Claude Code, and Gemini CLI. One pool, one router, one feedback loop — across all three hosts. Per-turn semantic top-K with dynamic context sizing, session-end self-evaluation, and evidence-blended re-ranking that gets better the more you use it.
CrossFrame Skill Suite:面向 AI agent 的中文结构诊断、公共议题分析、组织复盘、历史材料处理、追问和成文工作流。
Turn rough questions into concise prompt packs and safe Codex skill routes.
Local recommendation layer for routing Codex prompts to installed skills.
Universal AI skill router — auto-detects best installed skill per prompt + activates caveman mode for ~75% token reduction. Works with Claude, Codex, Cursor, OpenCode, Gemini CLI.
A Claude Code plugin that gives Claude a memory: it remembers your corrections, turns the repeated ones into principles, and invokes the right skill each turn.
AI Agent Semantic Routing Framework - Multi-layer semantic retrieval for intelligent skill dispatch
163 specialist Claude Code agents + 24 routing skills + 16 strategy playbooks. Korean/Japanese Business Navigators, game-dev (Unity/Unreal/Godot/Roblox/Blender), XR/spatial, paid-media routing. skill-routing-arbitrator disambiguates the ~500-skill ecosystem. Fork of msitarzewski/agency-agents.
Fix Gemma 4, Llama 4, Qwen 3, Mistral tool-calling failures. Zero-latency regex skill router for Ollama & OpenClaw — 400,000x faster than LLM inference, 100% accurate. No GPU needed.
Keyword-first skill routing and topology layer for file-based agent systems.
Self-growing Codex skill router that scans local skills, builds categories, recommends the best match in the user's language, and keeps learning from selections.
AI Agent Orchestration Framework — portable, project-agnostic workflow orchestration with checkpoint-based interruption and post-game review. Like a Makefile for AI workflows.
Architecture-first virtual task branching for agentic work
Adaptive skill routing for multi-agent systems — 4-dimension scoring engine
Local-first CLI to scan, audit, route, and test Codex Skills.
Memory and workflow automation for MCP, Codex, and Claude Code.
Local-first skill runtime + visual manager for local/open LLMs — source agentskills.io skills and inject the right one at runtime into LM Studio, MCP hosts, and any OpenAI-compatible proxy.
Kimi Code CLI fork: event-driven Orchestrator with pluggable policies, live SwarmProgressController showing each member's current tool call, /diag diagnostics, bundled Memory MCP server, minimax-backed web/image search chain. All opt-in, layered on top of upstream.
Add a description, image, and links to the skill-routing topic page so that developers can more easily learn about it.
To associate your repository with the skill-routing topic, visit your repo's landing page and select "manage topics."