Onboarding Guide

Anthropic Claude Compliance API

Connect Bloom Security to Anthropic’s Compliance API to get full visibility into Claude usage across your organization, including generated code, uploaded files, plugins, MCP servers, and configuration changes.

Why connect Anthropic Claude to Bloom?

Anthropic’s Compliance API gives you raw event data about what’s happening in your Claude environment. Bloom turns that data into actionable security intelligence by scanning content for secrets, monitoring your AI supply chain, and alerting on policy violations.

Secrets & Sensitive Data Detection

Automatically scan every artifact Claude generates and every file users share. Bloom detects hardcoded credentials, API keys, connection strings, and sensitive configurations before they become incidents.

AI Supply Chain Monitoring

Every new plugin, skill, and MCP server is a new component in your AI supply chain. Bloom tracks when they’re introduced, flags unreviewed components, and monitors lifecycle changes.

Configuration Monitoring

Get instant alerts when someone changes Claude’s execution policy or disables safety checks. Deploy guardrails with one click directly from the Bloom dashboard.

Before you start

Anthropic Claude Enterprise plan or Claude Console organization

Organization Primary Owner access in claude.ai

Bloom Security account with admin permissions

Getting started

1

Enable the Compliance API in Anthropic

1. Sign in to claude.ai as your organization’s Primary Owner. 2. Go to Organization settings > Data and privacy. 3. Click “Enable” under Compliance API.

2

Create a Compliance Access Key

In the same Data and privacy settings, click “Create key.” Grant the required compliance scopes, then copy the key—it will not be shown again.

read:compliance_activities

read:compliance_user_data

An Admin API key can access the Activity Feed, but cannot retrieve content. For full scanning, a Compliance Access Key is required.

3

Connect to Bloom Security

Log in to the Bloom Security dashboard. Go to Settings > Workflow Integrations > 3rd Party Integrations, paste your Compliance Access Key, then click “Test Connection” and “Enable.” Bloom begins ingesting events immediately.

4

Start monitoring

Navigate to any Claude software page in the Bloom dashboard—Claude Code, Claude Desktop, or Claude. Open the Activity tab to see real-time events with security scan results.

What you’ll see in Bloom

Claude Event

What Bloom does

Artifact created

Scans generated code for secrets and destructive commands.

File uploaded / accessed

Detects sensitive data shared with or accessed by Claude.

Plugin / Skill / MCP created or updated

Flags new AI supply chain components for review.

Execution policy changed

Alerts when organization policy is violated.

Need help getting set up?

Our team is here to help. Reach out to your Bloom Security account manager or contact support for a guided onboarding session.