xiachat-mcp_skill

This skill helps you discover compatible people and create personality profiles by matching SOUL.json using vector and LLM scoring.
  • Python

2.5k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill xiachat-mcp

  • _meta.json304 B
  • SKILL.md6.8 KB

Overview

This skill integrates XiaChat MCP features into MCP clients like Claude Desktop and Cursor, enabling AI personality matching, avatar pre-chat, and Soul Square persona conversations. It uses SOUL.json profiles and XiaChat API keys to find compatible matches, run automated pre-chats, and hand off to live conversations. The skill supports import/export with OpenClaw SOUL.md for cross-platform profile portability.

How this skill works

The skill inspects or generates SOUL.json personality profiles from names, quizzes, or chat logs and runs vector+LLM dual scoring to identify top matches. It can launch a 5-round AI pre-chat with a matched user to evaluate compatibility, produce a detailed report, and then hand off the live chat. It also provides browsing and conversational APIs for Soul Square personas and exposes credits/health checks for SOUL profiles.

When to use it

  • You want to find people whose personality profiles align with yours (dating, friendship, work).
  • You need to generate or convert SOUL profiles from chat history or OpenClaw SOUL.md.
  • You want an automated pre-chat to vet compatibility before a real conversation.
  • You want to explore or converse with curated AI personas from Soul Square.
  • You need to calculate or inspect compatibility scores between two SOUL profiles.

Best practices

  • Always set XIACHAT_API_KEY (xk_...) in your MCP server env before starting the service.
  • Create SOUL.json from both quiz answers and chat_text for richer embedding and LLM scoring.
  • Use match_find with an appropriate match_type (friend/dating/work) to bias results for your intent.
  • Run prechat_start on shortlisted matches to reduce time-investment before human handoff.
  • Export SOUL.md periodically if you maintain profiles across OpenClaw and XiaChat for consistent syncing.

Example use cases

  • Generate a SOUL.json from exported chat logs, then run match_find(top_n=5) to discover compatible friends.
  • Run prechat_start for a promising match; review prechat_report to decide whether to proceed.
  • Convert an OpenClaw SOUL.md into XiaChat SOUL.json to use XiaChat matching and Soul Square features.
  • Browse Soul Square personas (square_list) and start role conversations (square_chat) for creative brainstorming or coaching.
  • Compute match_score between two candidate profiles to support hiring or team pairing decisions.

FAQ

An XIACHAT_API_KEY in the format xk_... is required and must be set in the MCP server environment.

Can I interrupt the pre-chat early?

No. Pre-chat runs a fixed 5-round sequence and generates a report when complete; it cannot be stopped mid-run.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational