pa-code-mcp-server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@pa-code-mcp-server List all cubes and their dimensions in the model."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
pa-code-mcp-server
A standalone Model Context Protocol server for IBM Planning Analytics / TM1, extracted from the pa-code VS Code extension so it can run without the extension.
It exposes the TM1 model to any MCP-compatible LLM client as callable tools — both read (inspect cubes, dimensions, processes, rules, run MDX) and, in read/write mode, write / model-building (create dimensions, elements, consolidations, cubes, rules, TI processes, views, subsets; write cells; execute processes).
Local / proof of concept. Point read/write mode at a test database whose account has the rights the agent should have. Not a production, multi-user server.
Architecture
MCP client (Claude Desktop, …)
│ stdio or Streamable HTTP
▼
pa-code-mcp-server ──REST──> TM1 / Planning Analytics
(read + write tools)src/tm1/— standalone TM1 REST client + read/write API (no VS Code, no pa-code dependency).src/mcp/— MCP tool registration (mode-gated, confirm-guarded), server factory.src/index.ts— stdio transport (normal MCP launch).src/http.ts— Streamable HTTP transport (loopback, optional bearer token).
Related MCP server: ibm-mq-mcp
Setup
Requires Node.js ≥ 20.
cd C:\VSCode\pa-code-mcp-server
npm install
copy .env.example .env # then edit .env
npm run buildFill in .env:
TM1:
TM1_BASE_URL,TM1_AUTH_MODE(basic | cam | apikey),TM1_USER,TM1_PASSWORD, (TM1_CAM_NAMESPACE),TM1_SSL_REJECT_UNAUTHORIZED.Mode:
MCP_MODE=readonly(default) orreadwrite.
Run
As an MCP server (stdio) — e.g. from Claude Desktop or any MCP client:
npm startMCP client config example:
{ "mcpServers": { "tm1": { "command": "node", "args": ["C:/VSCode/pa-code-mcp-server/dist/index.js"], "env": { "MCP_MODE": "readwrite" } } } }As an HTTP MCP server:
npm run start:http # http://127.0.0.1:3900/mcpTools
Read (always): tm1_list_cubes, tm1_list_dimensions, tm1_list_processes, tm1_get_cube_dimensions, tm1_get_cube_rules, tm1_get_process_code, tm1_list_dimension_elements, tm1_list_hierarchies, tm1_list_views, tm1_list_subsets, tm1_execute_mdx, tm1_get_server_info.
Write (only in readwrite): tm1_create_dimension, tm1_create_hierarchy, tm1_add_elements, tm1_add_edges, tm1_delete_element, tm1_delete_dimension, tm1_create_cube, tm1_delete_cube, tm1_set_cube_rules, tm1_write_cell, tm1_write_cells, tm1_create_process, tm1_update_process, tm1_delete_process, tm1_execute_process, tm1_create_view, tm1_delete_view, tm1_create_subset, tm1_delete_subset.
Destructive/irreversible tools (deletes, cell writes, rule replacement, process execution) require a confirm argument equal to the exact target name.
Security notes
Read/write mode gives the agent the full TM1 rights of the configured account. Use a scoped test account.
TM1_SSL_REJECT_UNAUTHORIZED=falsedisables TLS validation (common for TM1 dev certs). Dev only.The HTTP transport binds to loopback; set
MCP_HTTP_TOKENto require a bearer token.
License
MIT — Copyright (c) 2026 Tim Geilen. See LICENSE.
IBM, Planning Analytics and TM1 are trademarks of IBM Corp.; this project is not affiliated with or endorsed by IBM.
Available Tools
13 toolstm1_execute_mdxA
Run a read-only MDX query (capped).
Inputs: mdx (required), maxCells.
tm1_get_cube_dimensionsC
Ordered dimensions of a cube.
Inputs: cubeName (required).
tm1_get_cube_gridA
Build a rows × columns value grid for an editable data-entry table. Returns JSON (rows, columns, values, updateable) that maps each cell back to a write tuple.
Inputs: cubeName (required), colElements (required), rowElements (required), colDimension (required), +4 more.
tm1_get_cube_rulesC
Rules text of a cube.
Inputs: cubeName (required).
tm1_get_process_codeC
Full source of a TI process.
Inputs: processName (required).
tm1_get_server_infoA
Static server configuration.
Inputs: no named parameters.
tm1_list_cubesC
List cubes.
Inputs: includeControl.
tm1_list_dimension_elementsB
List elements of a dimension hierarchy (capped).
Inputs: dimensionName (required), limit, hierarchyName.
tm1_list_dimensionsC
List dimensions.
Inputs: includeControl.
tm1_list_hierarchiesB
List hierarchies of a dimension.
Inputs: dimensionName (required).
tm1_list_processesB
List TI processes.
Inputs: includeControl.
tm1_list_subsetsC
List subsets of a dimension.
Inputs: dimensionName (required), hierarchyName.
tm1_list_viewsA
List views on a cube.
Inputs: cubeName (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 13 tools
Each tool maps to a distinct TM1 object or action: list_* enumerates resources, get_* retrieves specific detail, and execute_mdx/get_cube_grid cover querying versus grid-building. There is no meaningful overlap between tools.
All tools follow the consistent tm1_<verb>_<object> pattern. List is used for collection enumeration, get for detail retrieval, and execute for running a query, so the verb choice is predictable and uniform.
Thirteen tools is well within the ideal range for a domain-specific server. Each tool covers a meaningful part of browsing and querying a TM1 environment without unnecessary redundancy.
The read/exploration surface is strong: cubes, dimensions, hierarchies, processes, rules, MDX, and grid data are all covered. However, there are no write/update or process-execution tools, and the updateable grid has no corresponding save/commit tool, which is a notable gap for typical TM1 workflows.
Related MCP Connectors
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
An MCP server that provides an API to LLMs to manage their JumpCloud resources.
Governed data discovery, exact queries, decisions, simulations, and runtime utilities over MCP.
The Cortex MCP server provides read-only access to real-time engineering context from the Cortex developer portal, allowing AI coding assistants to answer natural language questions about your organization's catalog (microservices, libraries, domains, teams, infrastructure), scorecards (engineering standards and best practices), initiatives (goals and deadlines), and Engineering Intelligence metrics. It includes tools for querying documentation, tracking personal entities, and accessing AI-assisted insights across the entire Cortex ecosystem.
Related MCP Servers
- AlicenseAqualityCmaintenanceAn MCP server that connects AI assistants to Anaplan's Integration API v2, enabling users to browse workspaces, manage model data, and execute bulk operations like imports and exports. It provides 25 structured tools to navigate model hierarchies and perform transactional tasks through natural language.707MIT
- FlicenseAqualityBmaintenanceEnables LLM clients to interact with IBM MQ by exposing management and messaging REST APIs as 15 MCP tools, covering queue manager queries, MQSC commands, object inspection, and message send/receive operations.15-
- FlicenseNot gradedqualityAmaintenanceEnables LLMs to search for, add, configure, enable/disable, and aggregate tools from multiple MCP servers, dynamically extending their capabilities at runtime.-
- AlicenseNot gradedqualityAmaintenanceEnables LLMs to interact directly with generated REST APIs by exposing CRUD actions as MCP tools and read-only data as MCP resources.MIT