# pingfusi

> pingfusi — yutosolovyov770-pingfusi. Use this tool when you need to inject human oversight into AI-driven coding tasks, enabling real-time feedback and iteration to ensure accuracy and quality. It solves problems of unchecked AI output by inserting a review loop, where a human reviewer can pinpoint errors and provide a verdict. The tool integrates with git, accepting code inputs and producing refined, approved outputs.

Canonical page: https://skillsregistry.net/skills/yutosolovyov770-pingfusi  
JSON: https://api.skillsregistry.net/v1/skills/yutosolovyov770-pingfusi

## Description

MCP server + CLI that puts a real human in your coding agent's loop. It publishes work mid-task, a reviewer pins what's wrong and returns a verdict, and the agent iterates until approved.

## Trust

- **Trust score (0–1):** 0.50
- **Verification tier:** unverified

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** container
- **Runtime environment:** vm
- **Updated:** 2026-10-15

## Source

- **Source listing:** [GitHub](https://github.com/yutosolovyov770/pingfusi)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "yutosolovyov770-pingfusi"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/yutosolovyov770-pingfusi` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/yutosolovyov770-pingfusi/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt