agent-harness

Verify and bootstrap agent-readiness harness artefacts in software repositories.

4|Updated Mar 22, 2026
One-click install
npx skills add https://github.com/netresearch/agent-harness-skill --skill agent-harness-netresearch
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-harness
Source: https://github.com/netresearch/agent-harness-skill/tree/main/skills/agent-harness
Command: npx skills add https://github.com/netresearch/agent-harness-skill --skill agent-harness-netresearch

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Repositories often lack consistent, self-sustaining agent-readiness infrastructure, leading to unpredictable AI agent behavior, documentation drift, and repetitive manual setup work for every contributor regardless of their tools or skill installation status.

Core Features & Use Cases

  • Harness Verification: Run mechanical checks to validate AGENTS.md index format, documentation reference resolution, CI workflow presence, git hook configuration, and overall agent-readiness maturity level.
  • Harness Bootstrapping: Generate missing harness artefacts (AGENTS.md, CI workflows, git hooks, PR templates) from pre-built templates to quickly bring a repo to full agent-readiness.
  • Maturity Auditing: Assess a repo's harness maturity level (Basic, Verified, Enforced) and get actionable steps to progress to the next level. Use case: A team can run verification on every PR to catch documentation drift automatically, or bootstrap a new repository to get full agent enforcement in minutes without manual configuration.

Quick Start

Ask your AI agent to verify the harness consistency in the current repository to check its agent-readiness maturity level and identify any missing or inconsistent artefacts.

Frequently Asked Questions about agent-harness

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify AGENTS.md index format and documentation references in my repository?

To verify AGENTS.md index format and documentation references, run mechanical harness verification checks that validate agent-readiness artefacts, CI workflow presence, and git hook configuration. This process assesses your repo's maturity level to identify missing or inconsistent infrastructure components.

What is repository agent-readiness maturity and how do maturity levels work?

Repository agent-readiness maturity is a checkpoint-based model that measures infrastructure consistency across three levels: Basic, Verified, and Enforced. The maturity audit assesses your harness artefacts and provides actionable steps to progress to the next level of agent enforcement.

How do I bootstrap a new repository to get full agent enforcement with AGENTS.md and CI workflows?

To bootstrap a new repository for full agent enforcement, generate missing harness artefacts like AGENTS.md files, CI workflows, git hooks, and PR templates from pre-built templates. This quickly brings a repo to full agent-readiness without manual configuration.

Does agent-readiness harness verification work with GitLab CI and Forgejo CI systems?

Yes, agent-readiness harness verification works with GitLab CI and Forgejo CI systems, alongside GitHub Actions. The portable shell scripts integrate with these CI systems to run mechanical checks on every PR, catching documentation drift automatically across different contributor workflows.

What's the best way to automate agent-readiness checks on every pull request?

The best way to automate agent-readiness checks on every pull request is integrating portable shell scripts into your CI workflows. This enforces harness verification mechanically, validating AGENTS.md format, documentation references, and git hooks without requiring contributors to install specific tools.

Why does my repository have inconsistent AI agent behavior and documentation drift?

Inconsistent AI agent behavior and documentation drift occur because repositories lack self-sustaining agent-readiness infrastructure. Without mechanical harness verification and CI enforcement, agent behaviour becomes unpredictable and manual setup work is required for every contributor.