agent-routing

Sprix AI at 屿智同行 — state-aware SELF/COLLABORATE/HANDOFF routing for A2A agent networks.

Pre-inference tool and agent routing for MCP, A2A and tool-rich LLM systems. Reduce candidate tools before model inference with policy-aware routing and recovery.

Diverter distills Ultra’s delegation strategy into non-Ultra Codex sessions—bringing in the right native specialist while the main thread keeps moving.

Routes each request to the right AI agent or skill before your AI picks one. Your rules decide what code can; Jev (TypeSafe) makes the typed call. Installs as Claude Code hooks.

⚡ Sub-100ms cognitive reflexes for autonomous coding agents. Powered by TypeSafe AI's Jev & get-fable.

AI Agent Capability Router for Claude Code, Codex CLI, self-hosted LLM relays, OpenRouter-compatible gateways, and automation workflows.

High-Performance AI Agent Reverse Proxy & WAF Bypass Suite — Unicode homoglyph sanitizer, TLS header rotation & Claude CLI routing.

claude-node lets Python directly run, supervise, and integrate the real local Claude Code runtime. It preserves native CLI capabilities through stream-json, explicit session lifecycle, and process-level control — not another high-level agent framework, but a thin runtime layer for embedding Claude Code into your own systems.

Jev-LCT: Open System-One Decision Engine with Free Calibrated Confidence from Recurrent Trajectories

Markdown protocol for multi-vendor AI agent teams — capability cards, bandit routing, and lessons that compound.

Wald-Q4B: open-weight 4B decision model. Calibrated probability for every option, Jev-compatible /v1/systemone API, self-hosted. Weights on Hugging Face.

Open-source decision layer for AI agents. Typed task, tool, skill and review routing through TypeSafe Jev, with local policy gates and auditable receipts.

Astra-led orchestration for Codex: automatic model and thinking selection with 172 specialist agents.

Typed decisions from local open models. Python SDK, agent routing, and experimental Tetris. MIT.

Deterministic, auditable agent/skill dispatch for Claude Code. A typed seven-decision scoring kernel runs after the router has read the conversation — post-cognitive routing instead of prose-scanning. Auto-generates its catalog from skill sidecars and agent frontmatter.

Local-first LLM routing layer for agentic coding workflows, provider delegation, and cost-aware subagent execution.

Claude Code plugin marketplace from ogxo: cost-aware agent routing with a live board, multi-agent code review, git and PR workflows, browser debugging, guard hooks, a status line, and ThryX. The same plugins install in Grok Build.

Read-only Codex token auditing, adaptive model routing, and experimental Sol-Terra-Luna orchestration skills.

Stable reasoning-effort routing and privacy-safe decision metrics for Codex skills

Layered agent delegation that routes parallel subtasks across harnesses and models, then returns compact results to your planning agent.