How to fix perspectives not advancing?
Ensure `nextPerspectiveNeeded` is passed as a boolean `true`, not a string, and check server logs for debug output.
@FiveOhhWon
基本信息
MCP server for multi-perspective analysis and structured debates
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"perspective-mcp": {
"command": "npx",
"args": [
"perspective-mcp"
]
}
}
}Define the list of professional roles/personas that will analyze the topic
Perform analysis from the current perspective, building on previous insights
Start a debate between selected personas on a specific topic
Submit a statement in the ongoing debate
Add a new constraint or consideration to the ongoing debate
Generate a summary of the debate including key points from each participant
A Model Context Protocol server that enables AI assistants to analyze problems through multiple professional perspectives and facilitate structured debates between viewpoints. It integrates with MCP-compatible hosts like Claude Desktop, targeting developers, product managers, and decision-makers who need multi-stakeholder analysis.
Install via npx (npx perspective-mcp) or from source by cloning the repository, running npm install and npm run build. Configure the server in Claude Desktop's config file using either an absolute path to the built index.js or the npx command. Use the provided tools (set_perspectives, perspective, start_debate, debate_turn, inject_constraint, debate_summary) to set up professional roles, perform analyses, and conduct structured three-round debates.
Ensure `nextPerspectiveNeeded` is passed as a boolean `true`, not a string, and check server logs for debug output.
Verify all participant roles exist in your set perspectives and that 2-4 participants are selected.
No. It requires no file system access, makes no network requests, processes all data locally, and is stateless between sessions unless explicitly maintaining context.
Node.js 18 or higher and npm or yarn.
Yes, use the `inject_constraint` tool to add new considerations during the debate.
AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

B2B competitive intelligence, live sales battlecards, 5-axis ICP prospect fit scoring, and tech stack teardowns over MCP and REST with cited public-web evidence.
Hindsight: Agent Memory That Learns
不止于 CSS
YC (S26) | Open Computer History | Continuously record your company computer work, map your workflows, help you find work worth automating, and power your agents' context
评论