What is the Agent Council skill?

A skill that orchestrates a council of LLM agents to deliberate and vote on the best course of action for a given query.

What inputs does the Agent Council skill require?

query (string): The question or task for the council to deliberate on. agents (array): List of agent configurations, each with name, instructions, model.

What are the agent configurations in the Agent Council skill?

Each agent has: name (string), instructions (string): Role and behavior, model (string, default 'gpt-4o-mini').

What outputs does the Agent Council skill produce?

decision (string): The final council decision. reasoning (string): Summary of the deliberation process. votes (array): Each agent's vote and rationale.

How does the Agent Council skill deliberation process work?

1. Moderator poses the query. 2. Agents propose solutions. 3. Agents score and critique proposals. 4. Agents vote on the best proposal. 5. Moderator synthesizes the final decision.

What example council members are provided?

Advocate, Devil's Advocate, Researcher, Synthesizer.

OpenClaw skill

agent-council

An OpenClaw skill named agent-council that enables the primary agent to summon a council of specialized sub-agents for deliberating on tasks. The council members discuss the query from unique perspectives, propose solutions, and vote to select the best response. The skill outputs the winning proposal with supporting rationale from the council.

View repository Open SKILL.md

Security notice: review the SKILL.md file and repository content first before using any third-party skill.

Files

Review the files below to add this skill to your agents.

_meta.jsonView
SKILL.mdView

SKILL.md content

Unable to load SKILL.md content from source.

How this skill works

Instantiates each agent using OpenClaw's agent framework with the provided config
Runs a deliberation loop for the specified number of rounds
In each round of the loop, provides the current context (question + all previous messages) to every agent
In each round, each agent generates a response proposing their view or updating based on others
In each round, appends all responses to the shared context
In the final voting round, prompts agents to vote for the best proposal (their own or others')
Tallies votes and selects the proposal with the most votes (ties broken by first agent)
Outputs the winning decision and the full conversation log

When to use it

When assembling a council of agents with specified roles to deliberate on a query
When facilitating rounds of discussion, rebuttals, and voting among agents to reach consensus
When requiring a structured multi-agent process to generate a final council decision on complex topics

Best practices

Start with 3-5 council members to balance diversity and efficiency
Use diverse models across council members for varied viewpoints
Monitor token consumption for large councils
Define council members with distinct roles covering different perspectives
Set voting_threshold appropriately for the use case
Increase max_rounds for contentious topics
Always review the final vote and rationale before acting
Ensure each council member has a unique name
Enable logging to capture full debate transcripts when needed

Example use cases

Collaborative deliberation on complex problems: Assembles a council of AI agents with specific roles to discuss a given topic, propose solutions, vote, and reach a consensus decision.
Multi-agent decision making for strategic tasks: Enables specialized agents (e.g., CEO, CTO, CFO) to debate and decide on business or technical strategies like product launches.
Diverse perspective analysis: Leverages agents with unique expertise to provide multiple viewpoints, deliberate, and synthesize a final output on challenging topics.

FAQs

More similar skills to explore

achurch
An OpenClaw skill for church administration that handles member management, event scheduling, sermon retrieval, and donation processing. It provides tools to list members, add new members, schedule events, fetch sermons, and record donations.
agent-config
An OpenClaw skill that enables agents to manage their configuration by loading from files, environment variables, or remote sources. It supports retrieving, setting, and validating configuration values. The skill allows for hot-reloading of configurations.
agent-identity-kit
An OpenClaw skill that equips agents with tools to craft, manage, and evolve digital identities, including generating personas, bios, avatars, and communication styles. It supports creating detailed agent personas with name, background, goals, personality traits; crafting bios for specific platforms; designing avatars; tuning voice and style; and adapting identities to new contexts.
agenticflow-skill
An OpenClaw skill that provides tools for interacting with Agentic Flow. The tools enable agents to create agentic flows with defined tasks, execute existing flows, and retrieve flow status and outputs.
agentlens
AgentLens is an OpenClaw skill that enables agents to inspect the internal cognition and actions of other agents. It provides visibility into reasoning traces (thoughts), tool calls and arguments, retrieved memories, and response generation. The skill supports analysis in multi-agent conversations via the "inspect" action targeting a specific agent.
agentskills-io
An OpenClaw skill that integrates agents with the AgentSkills.io platform. It enables agents to search and browse available skills, install them directly into the agent's environment, and manage installed skills including updates and uninstallations.