# recon-sentinel

> recon-sentinel — alpha9934-recon-sentinel. Use this tool when you need to reconcile and troubleshoot issues in complex systems, particularly those involving language graphs and trust boundaries. It solves problems related to system inconsistencies and errors by providing a bounded state machine, trust-boundary tools, and human-approval interrupts. The recon-sentinel tool takes in system state inputs and outputs diagnostic reports and recommendations, making it ideal for use in development and testing contexts, such as git-based version control systems.

Canonical page: https://skillsregistry.net/skills/alpha9934-recon-sentinel  
JSON: https://api.skillsregistry.net/v1/skills/alpha9934-recon-sentinel

## Description

Agentic reconciliation-break triage system — bounded LangGraph state machine, MCP trust-boundary tool split, human-approval interrupt, and a real eval harness.

## Trust

- **Trust score (0–1):** 0.82
- **Verification tier:** verified
- **Last scanned:** 2026-09-28

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** container
- **Runtime environment:** vm
- **Updated:** 2026-09-28

## Source

- **Source listing:** [GitHub](https://github.com/alpha9934/recon-sentinel)

## 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": "alpha9934-recon-sentinel"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/alpha9934-recon-sentinel` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/alpha9934-recon-sentinel/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