MCP Server

The @phosra/mcp package exposes the Phosra API as MCP (Model Context Protocol) tools, allowing AI agents like Claude to manage child safety policies through natural conversation.

Installation

Run directly with npx:

bash

npx @phosra/mcp --api-key=phosra_live_sk_your_key_here

Or install globally:

bash

npm install -g @phosra/mcp
phosra-mcp --api-key=phosra_live_sk_your_key_here

Claude Desktop Configuration

Add this to your Claude Desktop claude_desktop_config.json:

json

{
  "mcpServers": {
    "phosra": {
      "command": "npx",
      "args": [
        "@phosra/mcp",
        "--api-key=phosra_live_sk_your_key_here"
      ]
    }
  }
}

On macOS, the config file is at ~/Library/Application Support/Claude/claude_desktop_config.json.

Available Tools

The MCP server exposes tools grouped by domain:

Families

Tool	Description
`list_families`	List all families for the authenticated user
`create_family`	Create a new family group
`get_family`	Get details for a specific family
`delete_family`	Delete a family and all associated data

Children

Tool	Description
`list_children`	List children in a family
`add_child`	Add a child to a family with name and birth date
`get_child`	Get a child's details and age group
`delete_child`	Remove a child from a family
`get_age_ratings`	Get age-appropriate content ratings for a child

Policies & Rules

Tool	Description
`list_policies`	List policies for a child
`create_policy`	Create a new policy for a child
`activate_policy`	Activate a draft or paused policy
`pause_policy`	Pause an active policy
`generate_rules`	Auto-generate age-appropriate rules for a policy
`list_rules`	List all rules in a policy
`update_rule`	Modify a specific rule's config or enabled state
`bulk_upsert_rules`	Create or update multiple rules at once

Enforcement

Tool	Description
`trigger_enforcement`	Push active policy to connected platforms
`get_enforcement_status`	Check the status of an enforcement job
`get_enforcement_results`	Get per-platform enforcement results
`retry_enforcement`	Retry a failed enforcement job

Platforms & Compliance

Tool	Description
`list_platforms`	Browse all supported platforms
`connect_platform`	Connect a platform to a family
`verify_connection`	Re-verify platform credentials
`disconnect_platform`	Remove a platform connection

Quick Setup

Tool	Description
`quick_setup`	One-call onboarding: create family, child, policy, and rules

Example Conversations

Setting Up a New Child

User: Set up parental controls for my 8-year-old daughter Maya.

Claude: I'll use Phosra to set up age-appropriate protections for Maya. Let me create everything in one step.

Calls quick_setup with child_name="Maya", birth_date="2017-06-15", strictness="recommended"

Done! I've set up Maya's protections:

Age group: 6-9 (Elementary)

Screen time: 90 minutes/day

Content rating: TV-Y7, PG movies, E for games

Web filter: Strict (safe search enforced)

45 rules configured across all categories

Adjusting Rules

User: Maya's teacher said she needs access to YouTube for homework. Can you allow that?

Calls list_rules to find the web_custom_allowlist rule Calls update_rule to add youtube.com to the allowlist

Done. I've added youtube.com to Maya's web allowlist. She'll be able to access YouTube while all other web filters remain in place.

Checking Enforcement

User: Did the rules push to NextDNS successfully?

Calls get_enforcement_status for the most recent job Calls get_enforcement_results to see per-platform details

Yes, enforcement completed successfully:

NextDNS: 12 rules applied, 0 failed

Last enforced 3 hours ago