War Room

Multi-agent war room for brainstorming, system design, architecture review, product specs, business strategy, or any complex problem. Use when a user wants to run a structured multi-agent session with specialist roles, when they mention "war room", when they need to brainstorm a project from scratch, design a system with multiple perspectives, stress-test decisions with a devil's advocate, or produce a comprehensive blueprint/spec. Works for software, hardware, content, business — any domain.

This skill appears internally consistent and low-risk, but exercise normal caution: (1) Run init_war_room.sh only in a workspace you control — the script creates files/folders under war-rooms/<project>. (2) Do not put secrets, API keys, or sensitive credentials in BRIEF.md, DNA.md, or any agent files (the workflow encourages free-text project input). (3) Review DECISIONS.md, agents/* outputs, and CHAOS challenges before acting on them — the system produces recommendations but they are not authoritative legal/technical approvals. (4) Be aware the skill is designed to spawn multiple subagent model invocations to simulate specialist roles; if you have policy limits on autonomous model calls, account for that. If you want a deeper assurance, provide the platform-specific execution sandboxing/permissions model or confirm where subagent invocations will run (same tenant/account or external).

Type: OpenClaw Skill Name: war-room Version: 1.1.0 The skill is classified as suspicious due to its use of high-risk capabilities, specifically the OpenClaw 'cron' API for scheduling tasks and the `open`/`xdg-open` commands for displaying artifacts (both mentioned in SKILL.md). While the skill includes explicit disclaimers stating these actions are within authorized platform APIs and scoped to the workspace, these functionalities represent significant control over system processes and external program execution. There is no clear evidence of intentional malicious behavior like data exfiltration or unauthorized remote control, but the presence of these risky capabilities, even for a stated purpose, prevents a 'benign' classification.