About P6XER MCP Server
P6XER MCP Server enables MCP-compatible AI clients to parse and analyze Primavera P6 XER schedule files. Built on PyP6XER, it provides tools, resources, and prompts for project activities, resources, schedule quality, and earned value.
Key features
- 01Parses XER project files
- 02Queries activities and critical paths
- 03Analyzes resource use and assignments
- 04Checks schedule quality
- 05Calculates earned value metrics
- 06Provides project resources and prompts
Use cases
- 01Summarize projects in an XER file
- 02Review critical path activities
- 03Check schedule quality and resource allocation
- 04Calculate project earned value metrics
MCP support
Exposes 13 tools, 3 resources, and 2 prompts for MCP clients to parse, query, and analyze .xer schedule files.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
uvx p6xer-mcp-serverP6XER_API_KEYSecretAPI key used to authenticate with p6xer service.P6XER_ENDPOINTOptional p6xer endpoint URL (default uses local p6xer daemon).P6XER_LOG_LEVELOptional log verbosity level (e.g., DEBUG, INFO, WARN, ERROR).Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
13Saved tool list · updated 3w ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| file_path | string | Yes |
Details
From the project READMEP6XER MCP Server
mcp-name: io.github.osama-ata/p6xer-mcp-server
A full-featured Model Context Protocol (MCP) server for Primavera P6 XER files, built on PyP6XER.
Exposes 13 Tools, 3 Resources, and 2 Prompts so any MCP-compatible AI client (Claude Desktop, Claude Code, Cursor, etc.) can interactively parse, query, and analyze .xer schedule files.
🚀 Features
🔧 Tools (13)
| Tool | Description |
|---|---|
parse_xer_file | Parse an XER file — project list, totals, status breakdowns |
get_project_activities | Activities with filters: project_id, project_short_name, status, task_type |
get_critical_path | Critical path activities (float ≤ 0), sorted by early start |
analyze_resource_utilization | Planned/actual hours & costs per resource; over-allocation flags |
check_schedule_quality | DCMA-style check: missing logic, long durations, high float, unresourced tasks |
get_resources | List resources, optionally filtered by type |
get_resource_assignments | Resource–activity assignments with enriched names and costs |
get_wbs | Work Breakdown Structure hierarchy |
get_relationships | Predecessor/successor relationships enriched with task codes |
get_calendars | Calendar definitions with hours-per-period data |
get_schedule_summary | At-a-glance stats: counts, date range, critical count |
get_earned_value | EVM: PV, EV, AC, CV, SV, CPI, SPI, EAC per project |
get_activity_detail | Full detail for one activity (preds + succs + resources) |
📋 Resources (3)
| URI | Description |
|---|---|
xer-project://{file_path}/{project_id} | Detailed text summary of a specific project |
xer-activities://{file_path} | Activities summary with status breakdown and duration stats |
xer-resources://{file_path} | Resources summary with type breakdown and assignment stats |
💬 Prompts (2)
| Prompt | Types |
|---|---|
analyze_xer_project | general · schedule · resources · progress · quality |
xer_reporting_prompt | executive · detailed · critical_path · resource · milestone |
� Install from Store
Via PyPI (uvx — no install needed)
uvx p6xer-mcp-server
Via PyPI (pip)
pip install p6xer-mcp-server
p6xer-mcp-server
Via Smithery
Search for p6xer-mcp-server on smithery.ai and click Install.
It will generate the correct Claude Desktop config automatically.
Via GitHub MCP Registry
The server is listed in the GitHub MCP Registry. In Claude Code:
claude mcp add p6xer -- uvx p6xer-mcp-server
Claude Desktop config (after PyPI install)
{
"mcpServers": {
"p6xer": {
"command": "uvx",
"args": ["p6xer-mcp-server"]
}
}
}
�📦 Installation
# Clone the repo
git clone https://github.com/osama-ata/p6xer-mcp-server.git
cd p6xer-mcp-server
# Install with uv (recommended)
uv sync
# Or with pip
pip install "mcp[cli]>=1.6.0,<2.0.0" pyp6xer
🏃 Running
Development / MCP Inspector:
uv run mcp dev src/p6xer_mcp_server/server.py
Stdio (Claude Desktop / Claude Code):
uv run p6xer-mcp-server
🔌 Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"p6xer": {
"command": "uvx",
"args": ["p6xer-mcp-server"]
}
}
}
Config file locations:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
🔌 Claude Code (CLI)
claude mcp add p6xer -- uvx p6xer-mcp-server
💬 Example Prompts
Once connected:
Parse project.xer and give me an overview of all projects
What are all the critical path activities, sorted by start date?
Run a DCMA schedule quality check on project.xer
Calculate earned value metrics — what are the CPI and SPI?
Show all resource assignments for "John Smith"
What are the predecessors and successors of activity A1000?
List all over-allocated labor resources
Generate an executive summary report for project.xer
📊 Tool Reference
Filter parameters (most tools accept)
project_id– numeric P6 project ID (e.g."1234")project_short_name– project short name string (e.g."PROJ1")
Status codes
TK_NotStart · TK_Active · TK_Complete
Task types
TT_Task · TT_Mile · TT_FinMile · TT_WBS
Resource types
RT_Labor · RT_Mat · RT_Equip
🗂️ Project Structure
p6xer-mcp-server/
├── src/
│ └── p6xer_mcp_server/
│ ├── __init__.py
│ └── server.py # All tools, resources, and prompts
├── manifest.json # MCPB bundle manifest
├── mcp.json # MCP registry metadata
├── smithery.yaml # Smithery registry configuration
├── pyproject.toml
└── README.md
📋 Requirements
- Python ≥ 3.10
mcp[cli] >= 1.6.0, < 2.0.0pyp6xer >= 1.16.0
📄 License
MIT
🚢 Publishing to PyPI (GitHub Actions + Trusted Publishing)
This repository includes a release workflow at .github/workflows/publish-to-pypi.yml based on the PyPA guide and pypa/gh-action-pypi-publish.
1) Configure Trusted Publishers
Create trusted publishers for this exact workflow file:
- PyPI: https://pypi.org/manage/account/publishing/
- TestPyPI: https://test.pypi.org/manage/account/publishing/
Use these values:
- Owner:
osama-ata - Repository:
p6xer-mcp-server - Workflow file:
publish-to-pypi.yml - Environment:
pypi(for PyPI),testpypi(for TestPyPI) - Project name:
p6xer-mcp-server
2) Create GitHub Environments
In repository settings, create two environments:
pypitestpypi
Recommended security setup:
- Require manual approval for
pypi - No approval required for
testpypi
3) Build locally (optional sanity check)
python -m pip install --upgrade build twine
python -m build
python -m twine check dist/*
4) Release flow
- Push to
main: publishes to TestPyPI - Push tag
v*(for examplev0.1.1): publishes to PyPI
git tag v0.1.1
git push origin v0.1.1
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.