Features / AI Integration
AI-Powered Enterprise Architecture — Connect Your AI Agent to Your Architecture Data
Albumi ships with an MCP server that lets AI agents — Claude, Cursor, or any MCP-compatible tool — work directly with your architecture data. Not through a chat widget. Through your actual API.
Your agent can pull your entire architecture, generate new entities from documents, validate schema and referential integrity, run 70+ automated architecture audits, and push changes back — all as reviewable Architecture Change Requests.
# Install and authenticate
npx @albumi/mcp login
# Add to Claude Code
claude mcp add albumi -- npx @albumi/mcp
# Then just talk to your agent:
"Pull our workspace. Run a full audit.
Show me all critical findings."
What Your AI Agent Can Do
Generate Architecture
Give your agent a description of your IT landscape — a wiki page, a spreadsheet export, an architecture document, or just plain text — and it generates a complete workspace: applications, integrations, data objects, business capabilities, IT components, and initiatives. With proper UUIDs, referential integrity, and lifecycle metadata.
"Here's our IT landscape doc. Create everything in Albumi — applications, their integrations, data objects, and map them to business capabilities."
Validate Consistency
Three-stage validation: JSON schema compliance, referential integrity (every ID resolves), and circular reference detection. Catches broken references, orphaned entities, invalid enums, and structural errors — before anything reaches your production workspace.
"Validate the workspace. Fix any issues you find, then validate again."
Audit Architecture — 70+ Checks
Automated architecture audit across 16 categories: structural integrity, data quality, lifecycle coherence, network analysis, portfolio health, compliance risk (GDPR, PCI, SOX), technology risk, initiative alignment, and more. Each finding comes with severity (critical/warning/info), affected entities, and context.
"Run a full audit. Show me all critical findings and suggest fixes."
Push as Change Request
Changes go back to Albumi as an Architecture Change Request (ACR) — a reviewable, approvable package. Nothing changes in your production workspace until a human reviews and approves it. Your agent proposes, your team disposes.
"Push the changes. Name the ACR 'Q2 Architecture Cleanup'."
Real Use Cases
Not hypothetical. These are workflows you can run today.
Zero-to-Architecture
You have no EA tool yet. You have wiki pages, spreadsheets, or just knowledge in people's heads. Feed your documents to the agent: "Here's our application list, here's our integration map, here are our business capabilities. Create everything in Albumi." The agent generates a complete workspace, validates it, runs the audit, and pushes it. You go from zero to a managed architecture in an afternoon — not in 3-6 months.
Architecture Health Check
Pull your workspace and run the 70+ check audit. The agent identifies: critical applications on end-of-life technology, integrations without authentication on sensitive data, circular dependencies, GDPR-relevant applications without PII mapping, portfolio imbalances, and lifecycle date violations. You get a prioritized findings report — not a generic "things to improve" list, but specific entities with specific problems.
Bulk Data Quality Fix
Your architecture data has gaps — missing descriptions, no criticality levels, empty lifecycle dates. Instead of clicking through 200 forms, tell the agent: "Fill in missing business criticality for all applications based on their capabilities and integration count. Add descriptions where missing." The agent reads the context, makes informed suggestions, and pushes the updated data for review.
Compliance Assessment
The audit engine checks GDPR, PCI-DSS, and SOX correlation: applications marked as GDPR-relevant but without PII data mapping, integrations carrying sensitive data without proper authentication, data objects with "restricted" classification flowing through unauthenticated channels. The agent surfaces these as critical findings with specific entities and remediation context.
Architecture Refactoring
You're consolidating 3 CRMs into 1. Tell the agent: "We're keeping Salesforce. Migrate all integrations from Legacy CRM and HubSpot to Salesforce. Mark Legacy CRM and HubSpot as 'Eliminate'. Create a 'CRM Consolidation' initiative and link all affected applications." The agent modifies the workspace, validates references, and pushes the plan as an ACR.
What the Audit Checks
70+ deterministic checks across 16 categories. No AI guessing — these are rules-based assessments on your actual architecture data.
Structural Integrity
Duplicate names, self-references, hierarchy depth, port direction
Data Quality
Missing criticality, fit levels, classifications, descriptions
Architecture Consistency
Cloud provider/hosting mismatches, protocol consistency
Lifecycle Coherence
Date order violations, EOL dependencies, phase-out status
Network Analysis
Integration hubs, circular dependencies, data flow consistency
Portfolio Health
TIME distribution, critical app risks, capability coverage
Compliance & Risk
GDPR/PCI/SOX correlation, auth security, data retention
Technology Risk
EOL components, vendor concentration, technology monoculture
Initiative Alignment
Unlinked initiatives, conflicting timelines, scope gaps
Organizational Coverage
Orphaned entities, missing ownership, org-capability alignment
Data Governance
Unclassified data, PII without controls, SoR conflicts
Redundancy & Complexity
Duplicate capabilities, over-connected systems, integration sprawl
How It Works
Connect
Install the MCP server (@albumi/mcp), authenticate with your Albumi account, and add it to your AI agent. One-time setup, takes 2 minutes.
Talk
Tell your agent what you need in plain language. Pull a workspace, generate entities from a document, run an audit, fix data quality — your agent calls the right tools automatically.
Review
Every change the agent makes is pushed as an Architecture Change Request. You review the diff, approve or reject. Nothing touches your production data without your explicit approval.
Apply
Approve the ACR in Albumi's UI. The changes merge into your workspace. Your architecture is updated, your diagrams reflect the new state, your impact analysis uses the latest data.
Supported AI Agents
Albumi's MCP server works with any agent that supports the Model Context Protocol. The most popular options:
Claude Code
Anthropic's CLI agent. Add the MCP server with one command. Full tool access, prompts, and resources.
Cursor
AI-powered code editor with MCP support. Add via settings.json. Works in the editor sidebar.
Any MCP Client
The server uses standard MCP stdio transport. Any agent that implements the MCP client protocol works out of the box.
AI-Assisted, Not AI-Dependent
The MCP integration is a power tool — not a requirement. Every feature available through AI is also available through Albumi's web UI. You can use AI for initial setup and bulk operations, then switch to the UI for day-to-day work. Or use AI for audits and the UI for data entry. Or never use AI at all. It's additive, not mandatory.
The AI integration is included on all plans — including the free plan. No add-on, no premium tier.
Let Your AI Agent Do the Heavy Lifting
Install the MCP server, connect your agent, and start managing your architecture with AI. Free for up to 50 applications.