Pilot Meeting Assistant Setup

Deploy a meeting assistant system with 3 agents. Use this skill when: 1. User wants to set up an automated meeting assistant system 2. User is configuring an...

This skill is internally consistent with its stated purpose, but verify a few things before proceeding: - Confirm pilotctl and clawhub are the official/expected binaries and obtain them from trusted sources; the skill will call these CLIs directly. - Review each pilot-* skill the setup installs (pilot-cron, pilot-webhook-bridge, pilot-receipt, pilot-stream-data, pilot-share, pilot-archive, pilot-task-router, pilot-alert, pilot-slack-bridge) to learn what credentials or network access they require (Slack tokens, webhook URLs, calendar API keys, etc.). The setup does not itself request those secrets but the installed components may need them. - Be aware that the handshake behavior auto-approves trust when both sides send a handshake; only perform handshakes with hosts you control or trust to avoid connecting to unknown agents. - The manifest is stored under ~/.pilot/setups — back up any existing files there if needed and inspect the manifest before starting agents. - Confirm network exposure and firewall rules for ports mentioned (1002 for agent-to-agent events, 443 for external webhooks), and restrict access as appropriate. - The skill bundle contains only documentation/instructions (no code); commands are executed only when you run them, so exercise the usual caution when copying/running CLI snippets. If you want higher assurance, ask the skill author or vendor for the exact requirements of each pilot-* component and for signed/official releases of pilotctl and clawhub.

Type: OpenClaw Skill Name: pilot-meeting-assistant-setup Version: 1.0.0 The skill bundle provides a legitimate orchestration workflow for deploying a multi-agent meeting assistant system. It uses framework-specific utilities (pilotctl and clawhub) to install dependencies, configure hostnames, and establish communication channels between agents as described in SKILL.md and README.md. No evidence of data exfiltration, unauthorized execution, or malicious prompt injection was found.