AI Business Hierarchies - Autonomous Agent Companies

Build and operate autonomous AI businesses using hierarchical agent systems. Implements folder-based company structures with CEO, Managers, Supervisors, and...

This skill largely does what it says (creates folders, spawns agent sessions, and schedules reports), but it has several operational and safety gaps. Before running anything: 1) Inspect the two scripts (setup-business.sh and setup-daily-reporting.sh) and the generated report script to understand exactly what they will do and where files are written. 2) Verify you have the necessary CLI tools (sessions_spawn, sessions_list, sessions_history, jq) and understand their permissions; the skill does not declare these dependencies. 3) Be aware the scripts append cron jobs to your user crontab — back up crontab first and check for duplicates; edit the cron file if you want different schedules. 4) The reporting script aggregates agent messages into files under ~/business/... — review it for sensitive content before enabling to avoid accidentally storing secrets or PII. 5) The skill references integrations and actions (LinkedIn, outreach, email sequences) but offers no credential/config guidance — do not provide API keys or account passwords unless you understand how agents will use them and where logs are stored. 6) Test in a sandbox or controlled account first (do not run on production or on accounts with live financial/legal privileges). 7) If you proceed, monitor spawned agents closely and have a plan to kill or remove them and to remove cron entries (crontab -e) if behavior is unexpected. If you want help auditing specific lines or hardening the scripts (avoid appending duplicate cron lines, limit report content, add confirmation prompts, require explicit credential config), I can walk through suggested edits.

Type: OpenClaw Skill Name: miknas-ai-business-hierarchies Version: 1.0.0 The skill bundle implements an autonomous agent hierarchy that utilizes high-risk capabilities such as crontab modification for persistence and automated session history scraping for reporting. The `setup-daily-reporting.sh` script creates a cron job that executes `sessions_history` on all active agents, potentially exposing sensitive data from various business departments (e.g., accounting, HR) into local markdown files. While these behaviors are aligned with the stated goal of 24/7 business automation, the broad access to session data and the use of system-level persistence mechanisms represent a significant security risk, although no evidence of intentional malice or data exfiltration was found.