# imhuso/cunzhi

告别AI提前终止烦恼，助力AI更加持久

Repository: https://github.com/imhuso/cunzhi
Canonical: https://ross.abutalabs.com/products/cunzhi
Language: Rust
License: MIT
License Family: permissive
Topics: ai, mcp
Last push: 2026-05-12T07:13:20+00:00

## Health v2 (maintenance only)
Score: 63/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 82, release rhythm 56, longevity 33
- inputs: {"age_days": 466, "days_push": 113, "days_rel": 299, "gap_med": 0, "n_releases_24m": 26}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1402, forks 164 (observed 2026-08-28T04:04:37.745375+00:00)

## What it is
Cunzhi (寸止) is an MCP tool that intercepts AI assistants when they try to end a conversation prematurely, popping up a dialog so users can continue interacting. It also provides per-project memory management for development preferences and semantic code search, built in Rust with Tauri.

## Use cases
- stop my AI assistant from ending conversations too early
- add a continue-dialog popup to Claude Desktop via MCP
- store project-specific coding preferences my AI can recall
- give my AI coding assistant semantic codebase search
- manage prompts and memory for MCP clients

## When to choose
- you use MCP clients like Claude Desktop and find the AI wraps up conversations prematurely
- you want per-project memory of dev conventions plus quick popup interaction
- you want a fast cross-platform install via Homebrew or release binaries

## When to avoid
- you don't use MCP-compatible AI clients
- you need a fully headless/automated pipeline with no interactive popups
- you only need plain prompt engineering without tooling

## Facets
- artifact type: cli-tool
- maturity: active
- function: mcp, chatbot, developer-tools, gui
- domain: developer-tools, large-language-models
- platform: windows, cross-platform, rust
- tags: mcp-server, ai-conversation, prompt-engineering, tauri, memory-management, ai-agents, macos, linux

## Member repositories
- imhuso/cunzhi (main) score 63

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:37.745375+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:38:58.915589+00:00, confidence not recorded.
  - readme: https://github.com/imhuso/cunzhi (fetched 2026-08-28T04:04:37.745375+00:00, sha 6924a7a5172e)
- Data as of 2026-08-30T08:39:29.467469+00:00.