agent-harness-generator

mcp
Security Audit
Warn
Health Warn
  • License รขโ‚ฌโ€ License: MIT
  • Description รขโ‚ฌโ€ Repository has a description
  • Active repo รขโ‚ฌโ€ Last push 0 days ago
  • Low visibility รขโ‚ฌโ€ Only 5 GitHub stars
Code Pass
  • Code scan รขโ‚ฌโ€ Scanned 12 files during light audit, no dangerous patterns found
Permissions Pass
  • Permissions รขโ‚ฌโ€ No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

๐Ÿ› ๏ธ The meta-harness for AI agents โ€” scaffold your own focused, branded agent harness with its own npx CLI, MCP server, memory, learning loop, and witness-signed releases. Works with Claude Code, Codex, pi.dev, Hermes, OpenClaw, and RVM (hardware-isolated).

README.md

agent-harness-generator

The meta-harness for AI agents โ€” a harness that builds other harnesses.

Like ruflo is the meta-harness for Claude, this is the meta-harness for AI agents themselves: a system whose job is to produce focused, vertical, branded agent harnesses that run on any host. Pick primitives, pick content, supply identity โ†’ ship a npm-publishable harness with your own npx <name> CLI, MCP server, memory, learning loop, and witness-signed releases.

npm โ€” coming soon
Status โ€” scaffold landed
License MIT

Claude Code
Codex
pi.dev
Hermes
OpenClaw
RVM

Rust + WASM
NAPI-RS
GCP-gated publish
Witness signed

One line: A meta-harness โ€” a marketplace plugin + CLI that scaffolds your own focused, vertical AI agent harnesses with their own npx <name> command, MCP server, memory, learning loop, and brand โ€” that run unchanged on Claude Code, Codex, pi.dev, Hermes, OpenClaw, and RVM.

What's a meta-harness? A harness is a runtime that orchestrates AI agents (memory + routing + hooks + MCP + claims). A meta-harness is the level above: a harness whose product is OTHER harnesses. agent-harness-generator emits self-contained, npm-publishable harnesses you OWN โ€” same kernel, your branding, your agents, your marketplace presence. The kernel updates flow to your harness via @ruflo/kernel npm peer; the content stays yours.

One paragraph: Ruflo bundles primitives (MCP server, hooks, memory bridge, swarm coordinator, intelligence pipeline, claims, routing) WITH opinionated content (60+ agents, 30+ skills, 33 plugins). agent-harness-generator factors those apart. You pick the primitives, pick the content, supply a name + brand, and out comes a brand-new npm-publishable harness with its own CLI, MCP registration, memory namespace, and marketplace identity โ€” running on the host of your choice.


Status

Scaffold landed. The Rust workspace, npm workspace, CI matrix, and GCP-gated publish pipeline are committed. Implementation work continues on a /loop-driven cadence. The 17 ADRs in docs/adrs/ define the design.

Layer Status Where
Kernel design Designed ADR-002, ADR-002a
Rust crate skeleton (7 subsystems) Scaffolded crates/kernel/
WASM bindings (wasm-bindgen) Scaffolded crates/kernel-wasm/
NAPI-RS bindings Scaffolded crates/kernel-napi/
@ruflo/kernel runtime resolver Scaffolded packages/kernel-js/
create-agent-harness CLI Stub packages/create-agent-harness/
CI (Rust + wasm + Node matrix) Wired .github/workflows/ci.yml
Publish pipeline (GCP Workload Identity Federation) Wired .github/workflows/publish.yml
Security (cargo-audit, cargo-deny, npm-audit, CodeQL) Wired .github/workflows/security.yml
Smoke test contract Wired scripts/smoke.mjs
Host adapters (Claude Code / Codex / pi.dev / Hermes) Iter 2+ ADR-004
Templates + composer Iter 2+ ADR-003

Architecture in 60 seconds โ€” the meta-harness pattern

   You                                  <- harness AUTHOR (uses the meta-harness)
       |
       v
   agent-harness-generator              <- THE META-HARNESS
   (this repo: scaffolds, signs, publishes harnesses)
       |
       v
   Your harness (npm package)           <- THE HARNESS YOU SHIP
       |
       v
   Your users
       |
       v
   npx <your-name>                      <- Identity (rename + brand)
       |
       v
   <your-harness>                       <- Content (your agents/skills/plugins/prompts)
       |
       v
   @ruflo/kernel                        <- Kernel (shared primitives, Rust + WASM + NAPI-RS)
       |
       v
   Host adapter                         <- Per-host abstraction
   (Claude Code / Codex / pi.dev / Hermes / OpenClaw / RVM)
       |
       v
   LLM providers

Read top-down: you (the harness author) operate the meta-harness. The meta-harness produces your harness. Your harness is what users install. They never see the meta-harness layer โ€” only the brand and CLI you ship.

The kernel is Rust source code compiled to two targets: WebAssembly (primary, cross-platform) and per-platform native binaries via NAPI-RS (escape hatch for hot Node paths). At load time, @ruflo/kernel prefers the native package for the current platform and falls back to wasm.

Working precedent: @ruvector/[email protected] โ€” 55 KB wasm-opt'd module shipping today through exactly this Rust โ†’ wasm-pack โ†’ npm pipeline.


Host support

Host Integration shape Notes
Claude Code MCP server + 5-handler-type hooks + 3-scope settings Ruflo-native target; richest hook surface
OpenAI Codex MCP via ~/.codex/config.toml [mcp_servers.*] tables TOML not JSON; no first-class hooks
pi.dev Pi extension (TypeScript via pi install npm:...) No MCP by design โ€” adapter uses pi.registerTool()
Hermes Agent MCP-supported runtime (optional-mcps/) Adapter scrubs <think> + stray <tool_call> per issue #741
OpenClaw MCP via ~/.openclaw/openclaw.json + workspace skills Personal AI assistant gateway with built-in multi-platform messaging (WhatsApp/Telegram/Slack/Discord)
RVM Bare-metal microhypervisor (AArch64) with capability tokens + hash-chained witness Hardware-isolated deployment for federated / multi-tenant / untrusted-peer scenarios (ADR-018)

See ADR-004 โ€” Host integration model.


Quality gates

Concern Where What it does
CI ci.yml Rust 3-platform matrix (fmt + clippy -D warnings + test + doc), wasm build + wasm-tools validate + 500 KB size budget, Node 20/22 ร— 3-platform tests
Publish gate publish.yml GCP Workload Identity Federation auth โ†’ Secret Manager fetches NPM_TOKEN โ†’ smoke test โ†’ npm publish --provenance (SLSA L2)
Security security.yml cargo-audit, cargo-deny, npm-audit, CodeQL, weekly cron
Smoke smoke.mjs Kernel loads, kernelInfo().version matches package.json, mcpValidate accepts/rejects correctly
Provenance ADR-011 Ed25519-signed witness manifest, byte-deterministic across CI runners (wasm enables this)

Quick start (developers)

git clone https://github.com/ruvnet/agent-harness-generator
cd agent-harness-generator

# Rust workspace
cargo test --workspace
cargo clippy --workspace --all-targets -- -D warnings

# WASM build
npm run build:wasm

# TypeScript + smoke
npm install
npm run build
npm run smoke

See CONTRIBUTING.md for the full developer workflow.


Read the design (17 ADRs + INDEX)

Start here: docs/adrs/INDEX.md

Highlights:


Related projects

  • ruflo โ€” the meta-harness this generator factors apart
  • ruvector โ€” vector + agentic database (memory backend)
  • @ruvector/emergent-time โ€” memory-decay clock the kernel uses
  • NAPI-RS โ€” Rust โ†’ Node bindings used for the native target

License

MIT โ€” see LICENSE.

Keywords: agent harness, agent harness generator, AI agent scaffolding, MCP server, Claude Code plugin, Codex plugin, pi.dev extension, hermes agent, multi-agent framework, agentic AI, agentic workflow, autonomous agents, agent orchestration, vertical AI harness, agent CLI generator, npm create agent, npx scaffold, Rust WASM kernel, NAPI-RS, wasm-bindgen, wasm-pack, agent memory, ReasoningBank, HNSW, emergent time, agent provenance, witness manifest, plugin marketplace, IPFS registry, drift detection, anti-slop, TDD, self-evolving agent, federated agents, swarm intelligence, GCP Workload Identity Federation, Secret Manager, SLSA provenance, npm provenance

Reviews (0)

No results found