Security audit
Security checks for vulnerabilities and agentic risk
The skill is purpose-aligned, but it asks users to run an unpinned external MCP package and persist an API key, so it should be reviewed before installation.
Install only if you trust the animalhouse.ai service and the npm package publisher. Prefer a pinned package version, treat the `ah_...` API key as a secret, and understand that the MCP server will contact a third-party service and can access the key you configure.
The skill instructs users to launch an external MCP server via npx -y mcp-animalhouse without pinning an exact package version or integrity mechanism. That creates a supply-chain risk: future package changes, compromise of the npm publisher, or dependency hijacking could cause arbitrary code to run in the user's environment when the MCP server is installed or executed.
The skill tells the agent to act on a broad natural-language request like “Register me at animalhouse.ai and adopt a duck named Quackers,” rather than requiring a narrow trigger and explicit confirmation for registration and external actions. In agent contexts, overly broad invocation guidance can lead to unintended account creation, credential handling, or network actions based on casual user phrasing.
The skill promotes creating an external account, issuing an API key, and persisting that key across sessions, but it does not prominently warn users about third-party data transmission, secret sensitivity, retention, or the risks of putting credentials into command history/configuration. That can cause accidental credential exposure or uninformed trust in an external service from within an agent workflow.
This second invocation repeats the same unpinned npx -y mcp-animalhouse guidance while also encouraging persistent API key configuration. If the package resolved by npm is modified or malicious, it could both execute arbitrary code and gain access to the configured ANIMALHOUSE_API_KEY, increasing the blast radius beyond a one-time install.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
No MCP? The same thing over HTTP:
curl -X POST https://animalhouse.ai/api/auth/register -H "Content-Type: application/json" -d '{"username": "your-name"}'
curl -X POST https://animalhouse.ai/api/house/adopt -H "Authorization: Bearer ah_your_token" -H "Content-Type: application/json" -d '{"name": "Quackers", "species_slug": "duck"}'
No suspicious patterns detected.