Team Discuss

What This Skill Does

Team Discuss is a powerful multi-agent collaboration framework designed to orchestrate complex group deliberation within the OpenClaw ecosystem. Unlike simple chatbots, this skill manages stateful discussions, tracks rounds of engagement, and enforces dialectical reasoning. By utilizing the 'sessions_spawn' mechanism, it allows you to bring different specialized agents together to debate, propose, and reach consensus on difficult topics. It acts as a moderator that prevents circular arguments by tracking previous messages and applying logic filters, ensuring that every participant contributes to a resolution rather than just stating opinions.

Installation

To integrate this skill into your environment, run the following command in your terminal:

clawhub install openclaw/skills/skills/chyher/team-discuss

Ensure that your environment has sufficient permissions for local file storage, as the skill uses a SharedStore to persist discussion states across agent sessions.

Use Cases

Team Discuss is best suited for scenarios involving high-stakes decision-making where multiple perspectives improve the outcome. This includes technical architecture reviews (e.g., microservices vs. monolith), product feature prioritization, ethical AI policy design, and academic research debates. It is particularly effective when you have specialized agents (e.g., a 'Backend Architect' and a 'Frontend Lead') that need to reconcile conflicting requirements.

Example Prompts

1. "Start a new discussion titled 'Database Migration' with the architect and devops agents, setting the max rounds to 5 and requiring a consensus threshold of 0.8."
2. "Run the 'API Versioning' debate and include a summary of any identified logical fallacies found in the architect's previous proposals."
3. "Summarize the current consensus on the 'UX Redesign' discussion and suggest an additional participant to break the current deadlock."

Tips & Limitations

To maximize the effectiveness of this skill, ensure your sub-agents have clearly defined roles and system prompts that emphasize cooperative problem solving. Because the skill relies on recursive calls, be mindful of the round limits to prevent excessive token consumption. The dialectical engine is a tool for transparency; while it flags potential fallacies, human oversight is recommended for critical business decisions. File-based persistence ensures data safety, but ensure your storage directory is backed up if managing high-volume, long-running simulations.