skills.rest
mifunedev
mifunedevOFFICIAL
Explore all skills by this author →
Website𝕏@mifunedevGitHub

harness-audit

Audit the Open Harness project from four perspectives and emit a tier-ranked report.

28|3|Updated Mar 27, 2026
Search Tags:#audit#agents#memory#wiki#ci-cd#git-worktree#crons
One-click install
npx skills add https://github.com/mifunedev/openharness --skill harness-audit-mifunedev
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: harness-audit
Source: https://github.com/mifunedev/openharness/tree/main/.claude/skills/harness-audit
Command: npx skills add https://github.com/mifunedev/openharness --skill harness-audit-mifunedev

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Spawn 4 parallel sub-agents (PM, Implementer, Critic, Explorer) to audit the harness for improvements. Synthesizes findings into tier-ranked actionable list. Outputs recommended next 3 actions.

Core Features & Use Cases

Run 4 parallel audit perspectives (PM, Implementer, Critic, Explorer), synthesize their ranked findings, and produce a single tier-classified improvement list with recommended next actions. Use-case examples include continuous health checks, onboarding reviews, and heartbeat audits to surface prioritized changes.

Quick Start

Ask the skill to run a four-perspective harness audit and return a tier-ranked action plan.

Frequently Asked Questions about harness-audit

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

FAQPage Schema
How do I run a multi-perspective audit on my CI-CD repository and workspace?▼

Run a four-perspective harness audit by spawning parallel PM, Implementer, Critic, and Explorer sub-agents that analyze repository artifacts, memory, wiki, and crons to produce a tier-ranked, actionable improvement list.

What is a harness audit and when should I use it for system health checks?▼

A harness audit evaluates project health across repository, workspace, and memory artifacts to surface prioritized improvements, making it ideal for onboarding reviews, continuous health checks, and heartbeat audits.

How do agents use memory and wiki artifacts to ground audit findings?▼

Agents ground their findings by reading structured frontmatter and repository artifacts like memory, wiki, and crons, ensuring the tier-ranked report reflects the actual project state rather than assumptions.

Can I use this audit approach for onboarding reviews of a git-worktree based project?▼

Yes, the harness audit supports onboarding reviews for git-worktree projects by evaluating repository artifacts, workspace memory, and wiki entries to surface prioritized next actions for new team members.

What is the best way to prioritize system improvements after a repository audit?▼

Synthesize multiple audit perspectives into a tier-classified improvement list that ranks findings by severity and outputs the recommended next three actions to execute immediately after the review.

Why does my parallel agent audit produce inconsistent improvement recommendations?▼

Inconsistent recommendations occur when structured frontmatter or repository artifacts like memory and wiki are missing or outdated, preventing sub-agents from grounding their findings for coherent tier-ranked synthesis.

Related Skills

Explore More Software Engineering Skills→
richfremrichfrem
5

gemini-cli-agent

Dispatch Gemini CLI sub-agents for persona-based analysis on large contexts.

Community
Intermediate
grant-vinegrant-vine
1

code-health

Audit code health across sources and tests into a severity-ranked Markdown report.

Community
Advanced
hollandkevinthollandkevint
5

git-workflow

Standardize Git branching, worktree management, and branch lifecycle closure.

Community
Intermediate
ZaxbyHubZaxbyHub
437

council

Coordinates three AI agents to jointly evaluate complex questions and produce cited advisory answers.

Official
Advanced