Audit OpenClaw harness configurations for health issues across workspaces.
npx skills add https://github.com/glfruit/gorin-skills --skill harness-audit-glfruit
Please help me install this Agent Skill. Skill: harness-audit Source: https://github.com/glfruit/gorin-skills/tree/main/openclaw/harness-audit Command: npx skills add https://github.com/glfruit/gorin-skills --skill harness-audit-glfruit
This Skill helps maintain healthy OpenClaw harness configurations by detecting bloated workspaces, stale references, duplicated rules, and hard-coded payloads across all agent workspaces.
Schedule regular scans with the harness-audit tool to obtain an up-to-date health report.
High-intent search queries and answers about installing and using this skill.
Checking SOUL.md and AGENTS.md identifies stale references and duplicated rules that degrade workspace health, using read-only checks that guide remediation without making prohibited changes without user confirmation.
Periodic health audits can be scheduled via cron to generate up-to-date reports, tracking trend data over time to identify configuration drift and improvement across all agent workspaces.
The OpenClaw audit tool performs non-destructive, read-only checks that guide remediation steps, strictly prohibiting configuration changes without explicit user confirmation.
The best way to track configuration drift is by running regular health audits that compare trend data over time, surfacing P0–P3 severity findings across all workspaces to identify drift and improvement.
You detect hard-coded payloads by running the harness-audit script to parse and report findings across SKILL.md files, classifying issues by severity to guide targeted remediation actions.
cyberpunk042Creates and registers new OpenClaw agents with metadata and workspace files.
kirvinAudit repository Markdown documentation and generate a prioritized findings report.
abzhawCoordinate OpenClaw agents, skills, and MCP servers with structured manifests.
foutoucourOrchestrate security assessments across code, infrastructure, auth, secrets, and pentest.
