# agent-safety

> Use this tool when you need to ensure the safe and responsible output of autonomous AI agents, scanning generated content before it is released to prevent potential harm or unintended consequences. It solves problems related to AI safety, mitigating risks associated with uncontrolled or malicious output. This tool accepts AI-generated output as input and returns a safety assessment, indicating whether the content is suitable for release.

Canonical page: https://skillsregistry.net/skills/compass-soul-agent-safety  
JSON: https://api.skillsregistry.net/v1/skills/compass-soul-agent-safety

## Description

Outbound safety for autonomous AI agents — scans YOUR output before it leaves the machine.

## Trust

- **Trust score (0–1):** 0.60
- **Verification tier:** unverified

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** instructions
- **Runtime environment:** llm
- **Category:** ai-ml
- **Updated:** 2026-05-17

## Source

- **Source listing:** [ClawHub](https://clawskills.sh/skills/compass-soul-agent-safety)

## 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": "compass-soul-agent-safety"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/compass-soul-agent-safety` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/compass-soul-agent-safety/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