DMtools Agent Skill
DMTools is an enterprise dark-factory orchestrator for project-level AI assistant workflows in Cursor, Claude, Codex, and other Agent Skills compatible systems.
Quick Install
IMPORTANT: This installer works with project-level directories only. Run it from your project root directory.
One-Command Installation (Recommended)
# Run from your project root directory
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/skill-install.sh | bashThe installer will:
- Detect project-level skill directories (.cursor/skills, .claude/skills, .codex/skills)
- Download the latest DMtools skill
- Automatically install to ALL detected locations (when piped)
- Install only in your current project (no global installation)
Example output:
Found skill directories:
1. .cursor/skills
2. .claude/skills
Non-interactive mode detected, installing to all detected locations
✓ Installed to .cursor/skills/dmtools
✓ Installed to .claude/skills/dmtoolsAdvanced Installation Options
# Interactive mode: choose specific location after downloading skill-install.sh
bash skill-install.sh
# Install to specific location only
INSTALL_LOCATION=1 bash skill-install.sh # First location (.cursor)
INSTALL_LOCATION=2 bash skill-install.sh # Second location (.claude)
# Install to all locations (explicit)
bash skill-install.sh --all
# Install only selected focused skills
DMTOOLS_SKILLS=jira,github bash skill-install.sh
bash skill-install.sh --skills jira,github
# Show help
bash skill-install.sh --helpInstall Only the Skills You Need
Use focused packages when you want dedicated slash commands such as /dmtools-jira or /dmtools-github instead of the full /dmtools bundle.
| Skill | Slash command | Latest package | Latest tarball |
|---|---|---|---|
| Full DMtools | /dmtools | https://github.com/epam/dm.ai/releases/latest/download/dmtools-skill.zip | https://github.com/epam/dm.ai/releases/latest/download/dmtools-skill.tar.gz |
| Jira | /dmtools-jira | https://github.com/epam/dm.ai/releases/latest/download/dmtools-jira-skill.zip | https://github.com/epam/dm.ai/releases/latest/download/dmtools-jira-skill.tar.gz |
| GitHub | /dmtools-github | https://github.com/epam/dm.ai/releases/latest/download/dmtools-github-skill.zip | https://github.com/epam/dm.ai/releases/latest/download/dmtools-github-skill.tar.gz |
| Azure DevOps | /dmtools-ado | https://github.com/epam/dm.ai/releases/latest/download/dmtools-ado-skill.zip | https://github.com/epam/dm.ai/releases/latest/download/dmtools-ado-skill.tar.gz |
| TestRail | /dmtools-testrail | https://github.com/epam/dm.ai/releases/latest/download/dmtools-testrail-skill.zip | https://github.com/epam/dm.ai/releases/latest/download/dmtools-testrail-skill.tar.gz |
Installer Selection Examples
# Install only Jira
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/skill-install.sh | bash -s -- --skills jira
# Install Jira + GitHub together
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/skill-install.sh | bash -s -- --skills jira,github
# Download the installer first, then reuse it locally
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/skill-install.sh -o skill-install.sh
bash skill-install.sh --skills ado,testrail# PowerShell installer usage
$env:DMTOOLS_SKILLS = "jira,github"
irm https://github.com/epam/dm.ai/releases/latest/download/skill-install.ps1 | iexManual Installation
- Download the latest
dmtools-skill-v<release>.zipfrom Releases - Extract to one of these project-level directories in your project root:
| Location | Platform | Description |
|---|---|---|
.cursor/skills/ | Cursor | Project-specific skill for Cursor IDE |
.claude/skills/ | Claude | Project-specific skill for Claude Desktop |
.codex/skills/ | Codex | Project-specific skill for Codex |
Note: The installer no longer supports global (user-level) installation. Each project should have its own skill installation to ensure version compatibility.
Child Agent Config Inheritance
When you need a base Agent Skill config plus specialized child configs, use the existing root-level parent block and deep-merge rules documented in Config inheritance via parent.
What’s Included
This skill provides comprehensive DMtools knowledge:
- Installation & Configuration - Setup guides for all platforms
- 67+ MCP Tools - Complete reference with examples
- JavaScript Agents - Development with GraalJS and MCP tools
- Test Generation - Xray, Cucumber, and automated testing
- Integrations - Jira, Azure DevOps, Figma, Confluence, Teams
- AI Providers - Gemini, OpenAI, Claude, DIAL, Ollama
Usage
Once installed, the skill is available in your AI assistant:
Cursor
- Type
/dmtoolsto invoke explicitly - Or mention DMtools and the agent will use the skill automatically
- View in: Settings → Rules → Agent Decides
Claude
- Type
/dmtoolsin chat - Or ask about DMtools naturally
Any AI Assistant
- The skill loads when relevant to your question
- Ask about installation, configuration, or development
Example Questions
- “How do I install DMtools?”
- “Help me configure Jira integration for DMtools”
- “Show me how to create a JavaScript agent with MCP tools”
- “Generate test cases from user story PROJ-123”
- “Set up Gemini AI provider for DMtools”
- “Create a teammate configuration for test generation”
Skill Structure
dmtools/
├── SKILL.md # Main skill definition (Agent Skills standard)
└── references/ # Detailed documentation
├── installation/ # Setup guides
├── configuration/ # Integration configs
├── agents/ # Agent development
├── test-generation/ # Test automation
└── examples/ # Code examplesUpdating
To update to the latest version, run from your project root directory:
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/skill-install.sh | bashNote: This will update the skill in your current project’s skill directories.
Compatibility
This skill follows the Agent Skills open standard and works with:
- Cursor IDE
- Claude Desktop
- Codex
- Any Agent Skills compatible system
License
Apache License 2.0
Resources
- DMtools Repository: github.com/epam/dm.ai
- Agent Skills Standard: agentskills.io
- Issues & Support: github.com/epam/dm.ai/issues
This skill enables AI assistants to provide comprehensive DMtools guidance and code generation.
All 63 pages
Installation
- DMtools Installation Guide
Back to README installation
- DMtools Installation Troubleshooting
Back to README installation
Configuration
- Azure DevOps (ADO) Configuration Guide
DMtools provides 23+ MCP tools for Azure DevOps integration, enabling work item management, sprint planning, and cross-platform synchronization with Jira.
- CLI Integration Configuration
The cli integration exposes cliexecutecommand — a tool for running shell commands from JavaScript agents and MCP workflows. It uses a security whitelist to prevent unintended command execution.
- CLI Output Formats
DMtools supports three output formats when executing MCP tools from the command line. The format controls how tool results are serialised and printed to stdout.
- DMtools Configuration Guide
DMtools uses a hierarchical configuration system with multiple sources. Configuration is loaded in priority order (highest to lowest):
- GitHub Configuration Guide
DMtools provides 8 MCP tools for GitHub integration, focused on pull request management: listing PRs, getting details, reading comments and review conversations, adding comments/labels, and getting…
- Google Gemini AI Configuration
Google Gemini provides powerful AI capabilities with a generous free tier (15 requests/minute), making it ideal for getting started with DMtools. Gemini 2.0 Flash offers excellent performance for m…
- Jira Configuration Guide
DMtools provides 52 MCP tools for Jira integration, enabling automated ticket management, test case generation, and workflow automation.
- JSON Configuration Rules for DMtools Jobs
The "name" field in DMtools job configuration is NOT a user-defined name or description.
- TestRail Configuration Guide
DMtools provides 15 MCP tools for TestRail integration, enabling automated test case creation, requirement linking, and test management directly from Jira/ADO stories.
MCP tools
- ADO MCP Tools
Create a new inline code comment on a specific file and line range in an Azure DevOps pull request. Creates a new thread with file context.
- Ai MCP Tools
- ANTHROPIC MCP Tools
Send a text message to Anthropic Claude AI and get response
- BEDROCK MCP Tools
Send a text message to AWS Bedrock AI and get response
- Bitbucket MCP Tools
- Bitrise MCP Tools
- CLI MCP Tools
Execute CLI commands from JavaScript post-actions. Base whitelist: git, gh, dmtools, npm, yarn, docker, kubectl, terraform, ansible, aws, gcloud, az. Additional commands can be enabled via CLIALLOW…
- CONFLUENCE MCP Tools
Get Confluence content by its unique content ID. Returns detailed content information including body, version, and metadata. Use format=md to convert body.storage.value to Markdown.
- Default Integration Aliases for MCP Tools
DMtools supports cross-integration aliases — universal tool names that map to platform-specific implementations. Instead of calling githublistprs or gitlablistmrs directly, you can call sourcecodel…
- DIAL MCP Tools
Send a text message to Dial AI and get response
- DMtools MCP Tools Reference — Overview
Complete reference for all MCP tools available in DMtools.
- DMtools MCP Tools Reference — TOOLS REFERENCE
This is the complete list of available MCP tools, generated directly from the running DMtools instance.
- FIGMA MCP Tools
Download image as file by node ID and format. Use this after figmageticons to download actual icon files.
- FILE MCP Tools
Delete file or directory from working directory. Returns success message or null on failure.
- GEMINI MCP Tools
Send a text message to Gemini AI and get response
- GITHUB MCP Tools
Create a new inline code review comment on a specific file and line in a GitHub pull request. To comment on a range of lines, provide both startLine and line. Side is 'RIGHT' for new code (default)…
- GITLAB MCP Tools
Create a new inline code review comment on a specific file and line in a GitLab merge request. Requires basesha, headsha, startsha from the MR diff refs (use gitlabgetmr to get them from diffrefs).
- Jenkins MCP Tools
- JIRA MCP Tools
Add a fix version to a Jira ticket (without removing existing ones)
- Jiraxray MCP Tools
- KB MCP Tools
Generate AI descriptions for existing KB structure WITHOUT processing new data. Use this after kbprocess to add AI-generated descriptions for people and topics.
- MERMAID MCP Tools
Generate Mermaid diagrams from content sources (Confluence or Jira) based on include/exclude patterns. Processes content recursively and stores diagrams in hierarchical file structure.
- OLLAMA MCP Tools
Send a text message to Ollama AI and get response
- OPENAI MCP Tools
Send a text message to OpenAI and get response
- Rally MCP Tools
- SHAREPOINT MCP Tools
Download a file from a SharePoint sharing URL to a local file path. Works with OneDrive and SharePoint sharing links.
- TEAMS MCP Tools
Complete device code authentication after user approval. Polls for tokens and saves refresh token.
- Teamsauth MCP Tools
- TESTRAIL MCP Tools
Create a new test case in TestRail with detailed fields (preconditions, steps, expected results, labels, type). Note: TestRail uses its own table format in text fields: |||:Col 1|:Col 2|:Col 3\n||v…
- VERTEX MCP Tools
Send a text message to Google Vertex AI Gemini (service account auth)
Jobs
- DMtools Jobs Reference
Complete reference for all 23 available jobs in DMtools. Jobs are specialized workflows that orchestrate MCP tools, AI agents, and data processing.
Agents
- AI Agent Development Best Practices
Critical patterns and lessons learned from real-world DMtools agent development.
- AI Teammate Configuration Guide
→ See also: Agent Best Practices for critical patterns, common helpers, and lessons learned from real-world development
- CLI Agent Integration with Teammate
→ See also: Agent Best Practices for critical patterns and common mistakes to avoid
- JavaScript Agents Development Guide
→ See also: Agent Best Practices for reusable helpers, common patterns, and critical preservation rules
Workflows
- GitHub Actions: AI Teammate Workflow
This workflow demonstrates how to run DMtools Teammate job in GitHub Actions for automated ticket processing.
- GitHub Repository Discoverability Playbook
Use this page as the maintainer source of truth for repository discoverability updates. The canonical metadata values are versioned in dmtools-core/src/main/resources/github-repository-discoverabil…
Test generation
- TestRail Manual Test Generation Guide
DMtools integrates with TestRail to automatically generate manual test cases from user stories (Jira/ADO). The TestCasesGenerator job creates test cases directly in TestRail, links them to source t…
- Xray Manual Test Generation Guide
DMtools integrates with Xray for Jira to automatically generate comprehensive manual test cases from user stories, including test steps, preconditions, and test sets.
Reporting
- Report Generator Skill Guide
This guide explains how to generate DMtools reports, which data sources and metrics you can use, and how to tune formulas, scores, and output. The example config for this guide lives at dmtools-ai-…
Skill packages
- dmtools-ado
dmtools-ado is the focused DMtools package for Azure DevOps work-item and pull-request workflows, including planning, ticket updates, comments, reviewers, and code-review thread management.
- dmtools-confluence
dmtools-confluence is the focused DMtools package for Confluence content discovery and page-management workflows, including page lookup, search, hierarchy traversal, and update automation.
- dmtools-figma
dmtools-figma is the focused DMtools package for Figma design-analysis workflows, including file structure inspection, node exports, image extraction, icon discovery, design-token access, team/proj…
- dmtools-github
dmtools-github is the focused DMtools package for GitHub workflows, including PR inspection, comment management, labels, review-thread follow-up, binary file attachment storage via GitHub Releases…
- dmtools-gitlab
dmtools-gitlab is the focused DMtools package for GitLab merge-request review workflows, including MR inspection, comments, review activities, and inline feedback.
- dmtools-jira
dmtools-jira is the focused DMtools package for Jira workflows, including ticket lookup, JQL search, comments, assignments, and Jira/Xray-driven delivery automation.
- dmtools-report
dmtools-report is the focused DMtools package for report-generation workflows, including configurable productivity reports, data-source aggregation, computed metrics, and HTML/JSON outputs.
- dmtools-teams
dmtools-teams is the focused DMtools package for Microsoft Teams communication workflows, including chat lookup, message retrieval, transcript discovery, file download, and sending messages.
- dmtools-testrail
dmtools-testrail is the focused DMtools package for TestRail test-management workflows, including case lookup, requirement linking, label handling, and generated test-case creation.
- Per-Skill Package Catalogue
This page is the canonical catalogue for the approved focused DMtools skill packages.