# Mcp Safety Review

> >

- **Type:** Skill
- **Install:** `agentstack add skill-arnie016-codex-prompt-templates-mcp-safety-review`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [Arnie016](https://agentstack.voostack.com/s/arnie016)
- **Installs:** 0
- **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [Arnie016](https://github.com/Arnie016)
- **Source:** https://github.com/Arnie016/codex-prompt-templates/tree/main/skills/mcp-safety-review
- **Website:** https://github.com/Arnie016/codex-prompt-templates

## Install

```sh
agentstack add skill-arnie016-codex-prompt-templates-mcp-safety-review
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# MCP Safety Review

Goal: decide whether a server/tool should be installed, sandboxed, modified, or rejected.

## Review Phases

1. **Source**: repo age, maintainer, license, release artifacts, install path.
2. **Runtime**: language, dependencies, postinstall hooks, Docker privileges, shell execution.
3. **Data**: files read, secrets touched, remote endpoints, telemetry.
4. **Tools**: classify every MCP tool by read/write/destructive/account-control capability.
5. **Auth**: API keys, OAuth, token storage, tenant boundaries, local-only claims.
6. **Blast radius**: what the tool can change without human confirmation.

## Severity

- `high`: credential exposure, destructive actions without confirmation, account takeover path, arbitrary command execution from untrusted input.
- `medium`: broad permissions, unclear auth boundary, unpinned remote code, weak logging of sensitive data.
- `low`: missing docs, weak error handling, optional hardening.
- `info`: posture notes without a concrete abuse path.

## Output

```md
# Safety Review
Verdict:
Severity:
Confidence:
Why it matters:
Evidence:
Required mitigations:
Install decision:
```

Hard rule: do not run raw binaries, zips, curl-piped scripts, or package
installers from a fresh repo until this review is complete.

## Pair With

- `$auto-skill-safety-mcp-secret-guard` when configs, logs, or examples may
  contain credentials.
- `$auto-skill-build-mcp-conformance-harness` after the server is trusted and
  the user needs repeatable CI/release gates.

## Validation

- Every exposed tool is classified by read/write/destructive/account-control capability.
- Install/run decision states what was not executed.
- Required mitigations are concrete enough to apply before trust is expanded.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [Arnie016](https://github.com/Arnie016)
- **Source:** [Arnie016/codex-prompt-templates](https://github.com/Arnie016/codex-prompt-templates)
- **License:** MIT
- **Homepage:** https://github.com/Arnie016/codex-prompt-templates

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** yes
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-arnie016-codex-prompt-templates-mcp-safety-review
- Seller: https://agentstack.voostack.com/s/arnie016
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.