Curated by
More in GitHub repos to check out
See all 54 →LinkedIn Engine Template for Development
github.com
Camelid: Rust-native Local Inference Backend
github.com
AI Agent Skill for Researching Topics Across Platforms
github.com
More from Mike Boscia
See all stacks →AI Inbox (Recovered)
15 cards · Technology
AI Brains
31 cards · Technology
Local AI & GPUs
36 cards · Technology
kkrlstrm/model-eval-gate: Versioned, eval-backed delegation policy for AI agents—prove what can leav
model-eval-gate is a delegation policy engine for AI agents that transforms observed work and evaluations into versioned policies to control model downgrades in multi-agent workflows. It sits between agent runtimes and gateways to enforce whether work can be delegated to smaller models while preventing unevaluated downgrades from becoming production behavior.
Built for AI agentsACO · 7161 tokens
Summary
model-eval-gate is a delegation policy engine for AI agents that transforms observed work and evaluations into versioned policies to control model downgrades in multi-agent workflows. It sits between agent runtimes and gateways to enforce whether work can be delegated to smaller models while preventing unevaluated downgrades from becoming production behavior.
Tags
ai-agents · delegation-policy · model-evaluation · quality-enforcement · agent-runtime · cost-optimization · policy-engine
Key entities
model-eval-gate (technology, 0.99) · OpenClaw (technology, 0.95) · OpenRouter (technology, 0.9) · LiteLLM (technology, 0.9) · Hermes (technology, 0.85) · delegation-policy (concept, 0.95) · quality-enforcement (concept, 0.9) · model-downgrade-control (concept, 0.88) · kkrlstrm (organization, 0.85)
Classification
reference · language en · status final
Provenance
claude-haiku-4-5 via @stacklist/be@0.1.0, confidence 0.85, 3 Sep 2026