Consulting Agent Pipeline

• Verdict: Pass
• Confidence: High
• Analyzed: Apr 15, 2026, 1:16 PM
• Findings: 0
• Version: 1.0.0

Overview

The skill's files and runtime instructions are coherent with its stated purpose (a filesystem-based multi‑Agent consulting pipeline); it does not request external credentials or network endpoints and contains only local helper scripts and templates.

This skill appears internally consistent and implements a filesystem-based multi‑agent pipeline with local helper scripts (scanning, validation, snapshot). Before installing/running: 1) Review and, if needed, edit templates/FORBIDDEN_TERMS.yaml so the hard-coded defaults match your project; 2) Be aware forbidden-terms-scan.sh will search project files and — if you don't provide a FORBIDDEN_TERMS.yaml path — will also probe a few $HOME locations (e.g., ~/.openclaw/... and ~/Documents/Obsidian Vault). If you keep sensitive data out of the shared project directory and avoid using cloud-synced home folders as targets, risk is low; 3) If you use Obsidian or any synced folder, remove or change the example Obsidian copy paths in templates to avoid accidental syncing; 4) Inspect and (optionally) run the scripts locally in a controlled project directory to confirm tool availability (yq, rsync, unzip) and behavior before letting any agent run them autonomously; 5) Do not use this skill for highly confidential projects (the SKILL.md already warns against that).