# Harness

A meta-skill that designs domain-specific agent teams, defines specialized agents, and generates the skills they use.

Repository: https://github.com/revfactory/harness
Canonical: https://ross.abutalabs.com/products/harness
Homepage: https://revfactory.github.io/harness/
Language: HTML
License: Apache-2.0
License Family: permissive
Topics: claude-code, claude-code-plugin, harness, harness-engineering
Last push: 2026-07-24T23:35:08+00:00
Link (homepage): https://revfactory.github.io/harness/

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 94, release rhythm 35, longevity 11
- inputs: {"age_days": 160, "days_push": 40, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 8838, forks 1251 (observed 2026-08-28T04:10:25.977045+00:00)

## What it is
Harness is a Claude Code plugin that acts as a meta-skill for designing domain-specific agent teams. From a single prompt, it generates agent definitions and skills using six team-architecture patterns such as Pipeline, Fan-out/Fan-in, and Supervisor.

## Use cases
- build a multi-agent team for deep research
- generate specialized claude code agents for my project
- orchestrate a full-stack website development pipeline with agents
- create a producer-reviewer agent team for content production
- auto-generate skills for claude code agents
- set up a supervisor pattern for multi-agent task delegation

## When to choose
- you use Claude Code and want to decompose complex projects into coordinated specialized agents
- you want battle-tested multi-agent patterns (pipeline, fan-out, supervisor) without designing them yourself
- you need auto-generated skills with progressive disclosure to keep agent context lean

## When to avoid
- you don't use Claude Code or its agent team system
- you need a deterministic, production-grade multi-agent runtime rather than generated team configurations
- you want a framework with runtime orchestration code instead of generated agent/skill definitions

## Facets
- artifact type: plugin
- maturity: active
- function: agent-framework, prompt-engineering, workflow-automation, developer-tools
- domain: large-language-models, developer-tools
- platform: cli, cross-platform
- tags: claude-code, claude-code-plugin, multi-agent-orchestration, meta-skill, agent-teams, skill-generation, ai-agents, automation, nodejs

## Member repositories
- revfactory/harness (main) score 57
- revfactory/harness-100 (examples) score 47

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:25.977045+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:24:48.683934+00:00, confidence not recorded.
  - readme: https://github.com/revfactory/harness (fetched 2026-08-28T04:10:25.977045+00:00, sha cae581651fab)
  - homepage: https://revfactory.github.io/harness/ (fetched 2026-08-29T08:25:32.819370+00:00, sha bba86c7f36a7)
- Data as of 2026-08-30T08:39:29.467469+00:00.