MCP Integration

Connect Claude Desktop & Claude Code to OrgGen

Your agent doesn't just edit a diagram: it creates tasks, raises a decision for the team to arbitrate, saves a named version, and runs your cap table — from Claude Desktop or Claude Code, without switching tools.

What is MCP?

The Model Context Protocol (MCP) is an open standard that allows AI assistants like Claude to securely connect to external tools and data sources. With OrgGen's MCP integration, Claude can directly read and modify your organizational schemas.

What You Can Do

📦

Tasks & arbitration

Your agent creates tasks and raises a decision for the team — a type-level lock structurally prevents the agent from validating it itself.

🧠

Named versions

Your agent saves a named version at each milestone — the one your partners then sign in the app: an immutable audit trail, a PDF attestation with a SHA-256 fingerprint.

💬

Cap table, live

Model a round, an exit, a SAFE conversion — the effect is computed and shown before anything is written.

📊

Multi-project cascade

Several linked projects held as one graph — your agent switches and compares, you keep the overview.

Setup Guide

1. Create an API Key

Go to your Settings page and create a new API key. Copy it immediately - it won't be shown again.

Go to Settings →

2. Configure Claude Desktop

Add the following to your Claude Desktop configuration file:

Config file location:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

• Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "orggen": {
      "url": "https://www.orggen.ai/api/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

3. Restart Claude Desktop

Restart Claude Desktop to load the new MCP server. You'll see OrgGen tools available in your conversations.

Using with Claude Code

For Claude Code, you can use the CLI quick install command:

Quick Install (CLI)

Run this command in your terminal:

claude mcp add --transport http orggen https://www.orggen.ai/api/mcp --header "Authorization: Bearer YOUR_API_KEY"

Or add manually to .mcp.json:

{
  "mcpServers": {
    "orggen": {
      "url": "https://www.orggen.ai/api/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Available Tools

Once connected, Claude gets access to 145 MCP tools, grouped below by domain. For the exact, up-to-date list your key actually sees, call the describe_toolset tool from your client.

Org chart, people & search

41
get_schema_overview, get_cell_details, search_cells, search_people, …and more

Data model (MCD/ERD)

17
add_entity, update_entity, delete_entity, add_entity_with_fields, …and more

Comments & mentions

6
list_comments, add_comment, reply_comment, resolve_comment, …and more

Data sources & lineage

6
add_data_source, update_data_source, delete_data_source, link_field_lineage, …and more

Data model export/import

3
list_data_model_formats, export_data_model, import_data_model

Generic multi-kind tools (arch/process/journey/cap table)

4
add_kind_node, update_kind_metadata, connect_kind_nodes, compute_group_ownership

Architecture (C4) export/import

3
list_arch_formats, export_arch_diagram, import_arch_diagram

Diagram export

2
export_diagram, export_fundraising_package

Architecture (C4) validation

1
validate_arch

Business process

3
validate_process_flow, add_swimlane, auto_layout_process

Customer journey

4
validate_journey, compute_emotion_curve, add_journey_phase, auto_layout_journey

Cap table & simulation

5
validate_captable, compute_direct_ownership, simulate_round, simulate_exit, …and more

Per-kind playbook

1
get_playbook

Tasks & arbitration

5
list_tasks, create_tasks, update_task, archive_task, …and more

Change proposals (no apply)

3
propose_changes, list_proposals, get_proposal

Versions & signed workflow

3
save_version, get_version, diff_versions

Org chart import

1
import_org

Cap table import (proposal)

1
propose_captable_import

Cross-project linking

2
link_project_to_cell, unlink_project_from_cell

Content block move

1
move_content_block

Documents attached to blocks

7
list_documents, get_document, create_document, update_document, …and more

Nested schema under a block

1
create_nested_project

Fork of a linked project

1
fork_linked_project

Reading a linked project

1
get_linked_project

Custom project metadata

3
set_project_metadata, get_project_metadata, delete_project_metadata

Moving a task to another project

1
move_task_to_project

Reading product feedback

1
list_feedbacks

Block identifier aliases (non-destructive rename)

1
add_cell_id_alias

Activity log (history)

1
list_activity

Child API key emission

1
create_api_key

Document revision history

3
list_document_revisions, get_document_revision, restore_document_revision

Tracking honesty (tasks & arbitrations)

1
analyze_tasks

Register of decisions in force

1
list_decisions

Bulk task import (replayable, reversible)

3
import_tasks, rollback_import, list_imports

Atomic task claim by an agent

1
claim_task

Claim renewal and declared progress

1
renew_claim

Consolidated report of what changed since a cursor

1
get_contextual_updates

Reading the project’s rules (configuration)

1
get_config

Revoking a child API key from its parent key

1
revoke_api_key

Block roadmap: read the commitment horizon, suggest a horizon

2
list_roadmap, suggest_horizon

Frequently Asked Questions

Is MCP access paid?

No. Access to OrgGen's MCP server is included in the Free plan — just create an account and an API key, no payment required. It's rate-limited to 500 calls/day; paid plans raise that limit.

Which MCP clients are supported?

Any agent or client that speaks MCP — the Model Context Protocol — over the streamable-http transport.

How many tools does the MCP server expose?

145 MCP tools across 40 domains (org chart, data model, C4 architecture, process, customer journey, cap table, tasks, documents, comments, versions…). Call describe_toolset for the exact, current list for your key.

Can the MCP server modify my data, or only read it?

Both, depending on the API key's scopes (schema:read, schema:write, feedback:read, feedback:write). Every write call is logged, versioned, and reversible through version history.

🔒 Security

Your API keys are hashed and stored securely. Each key has specific permissions (scopes) that control what actions Claude can perform. You can revoke keys at any time from your Settings.

Ready to Get Started?

Create your first API key — your agent gets 145 tools, from schema to signature.

Create API Key