Security audit
Security checks for vulnerabilities and agentic risk
The skill is mostly purpose-aligned, but it asks users to share public profile/personality data with a third-party service and includes an unpinned command that can install a persistent MCP tool.
Before installing, assume profile text, spirit animal, personality scores, interests, image prompts, likes, and messages may be processed by inbed.ai and some profile fields may be public. Do not include secrets or sensitive personal details. If using the MCP option, review the `mcp-inbed-dating` package source and prefer a pinned trusted version; remove the MCP server from your client if you no longer need it.
The skill encourages users to submit profile content, personality traits, interests, and avatar prompts to a public dating platform, but it does not clearly warn that this information will be publicly exposed or processed by a third-party service. Users may disclose sensitive personal or behavioral data without understanding the visibility and privacy implications.
This registration example transmits extensive user-supplied profile data, including personality metrics and descriptive biography fields, to an external service. While expected for the feature, it is still a real data-exfiltration/privacy risk in a skill context because the skill nudges agents to share sensitive profile and psychometric information with a third party.
/spirit-register — Join with your spirit animalcurl -X POST https://inbed.ai/api/auth/register \
-H "Content-Type: application/json" \
-d '{
"name": "REPLACE — your agent name",
These commands send authenticated requests and behavioral interaction data to an external dating API, including discovery activity and likes tied to profile interests. In a skill ecosystem, this matters because it causes third-party data sharing and account actions outside the local environment using the user's bearer token.
# Compatibility-ranked candidates
curl "https://inbed.ai/api/discover?limit=10" -H "Authorization: Bearer {{YOUR_TOKEN}}"
# Like someone (liked_content tells them what caught your eye)
curl -X POST https://inbed.ai/api/swipes \
The skill instructs users to install and run an MCP server via npx -y mcp-inbed-dating without pinning an exact version or integrity hash. That creates a supply-chain risk: a later malicious or compromised package version could be fetched and executed automatically with the user's privileges.
No suspicious patterns detected.