MCP App StoreMCP App Store
from claw-router18
Trait-based router that scores messages by complexity and task type to automatically select the most suitable AI model for each request.
Claw Router scores incoming messages across complexity tiers and task-type dimensions, then matches those traits to configured model profiles to route each request to the best-fit AI model. It supports rule-based scoring augmented by LLM arbitration for borderline cases or tie-breaking.
Use this skill when an environment hosts multiple LLMs and you need dynamic routing to optimize cost, latency, or capability—for example, sending simple chat queries to a low-cost model while routing complex coding or design tasks to a stronger model. Also useful for hybrid deployments where models specialize in translation, math, writing, or analysis.
Inferred compatibility: OpenClaw agents, orchestration plugins, or any system that can call a routing tool; integrates well with Claude Code, Copilot-like agents, or custom LLM services for arbitration.
Claw Router is a trait-based model routing skill that scores incoming messages by complexity tier and task type, then matches against user-declared model trait profiles to select the optimal AI model. The SKILL.md is clear and well-structured with good configuration examples and a fixed trait vocabulary. However, the LLM assist section includes a hardcoded apiKey placeholder in the config example, and the skill references an external API endpoint (api.siliconflow.cn) for optional LLM scoring, which poses a phone-home risk. No scripts were present to test.
Security deductions: -18 for telemetry/phone-home to external host (LLM scoring sends messages to user-configured API), -12 for the apiKey placeholder in config example which encourages pasting real keys into config, -12 for unvalidated external API endpoint risk. The skill is a config-driven router concept — useful for multi-model setups but the actual routing engine isn't bundled in the skill itself, so it's more of a specification than a complete skill.
Yielding Bear — Unified LLM Routing
A unified LLM routing API that automatically routes requests to the cheapest capable model across providers to reduce inference cost by ~60-80%.
Starlark Dev
Create, debug, and test Kurtosis Starlark packages — write package structure, run dry-runs, and inspect plan execution for reliable orchestration.
Maestro Orchestrate
Orchestrates complex engineering workflows with enforced design and approval gates, child-agent delegation, and session state management.
Run Execute
Orchestrates execution of work items across modes (autopilot, confirm, validate) with scripted init/complete tooling, plan/test/report artifacts and strict gati
Overnight — Autonomous Long-Running Coding
Orchestrates long-running coding goals: decomposes objectives into atomic tasks, dispatches isolated worktree workers, verifies acceptance criteria, and merges
tmux Remote Control
Control tmux sessions for interactive TTYs: send keystrokes, capture panes, orchestrate parallel coding agents, and monitor output safely.
Build Unplanned Feature
Orchestrates ad-hoc feature development: requirement gathering, planning, implementation and validation across task-planning, implementation, and validation loo
AI Pair Collaboration
Orchestrates a small heterogeneous team of AI agents (author/developer + Codex reviewer + Gemini reviewer) to produce and review content or code in a coordinate