The agentic meta-harness — freeze the model, evolve the harness. An open runtime that routes each query to the cost-optimal model, evolves its own harness (planner/context/reviewer/retry/tool/memory/score policy) and autonomously repairs code, then orches

Track E / §56 — Meta-Harness-style optimizer (arxiv 2603.28052). Coding-agent proposer with filesystem-backed full history. BREAKING: opt-in mutator that can rewrite the bundle directly, sidestepping the spec-as-source-of-truth invariant.

Meta-harness for Kiro — agent orchestration, persistent memory, task coordination, and self-learning capabilities

A meta-harness for designing durable loops that prompt coding agents

Loom — A manifest-driven agent meta-harness

CrewHaus — the meta-harness compiler for AI agents. Compile one crewhaus.yaml spec into a CLI agent, channel bot, RAG pipeline, multi-agent crew, eval harness, voice or browser agent, and more.

Auto-tuned harnesses for AI agents — Meta-Harness algorithm with Pareto search over (accuracy, tokens, latency)

AgentRig — an agentic meta-harness. A CLI that investigates a repository and installs (and evaluates) a best-practice agent harness.

Viterbi — an agnostic meta-harness for agentic engineering. Installs a global rule + memory surface, primitive skills, and a full project scaffold across AI coding runtimes (Claude Code, Codex, and more).

Harbor — The Skills Operating System for Claude Code. Automatic skill detection, subagent swarms, meta-harness security, and 90%+ token reduction.

Spec authoring + inline trace inspection + run-from-editor for the CrewHaus meta-harness compiler (Section 35).

MRAgent — Cue-Tag-Content graph memory over RuVector, with a Meta-Harness Darwin loop that evolves the reconstruction harness (freeze the model, evolve the harness).

Authority-bound installed execution custody for approved one-shot repository changes.

Browser-based REPL for the CrewHaus meta-harness compiler — Monaco editor + live trace timeline + OAuth quota + cross-tenant isolation (Section 35).

TypeScript SDK for Strands Agents framework

FH runtime adapters — run FH governance, skills, and agents via Claude or Codex with machine-parseable gates.



The core test runtime that executes on React Native devices, providing Jest-compatible APIs (describe, it, expect) and managing test collection, execution, and result reporting in native environments.