Enterprise dark-factory orchestrator for automating delivery workflows across trackers, source control, documentation, design systems, AI providers, and CI/CD.
Quick install
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/install.sh | bashPowerShell: https://github.com/epam/dm.ai/releases/latest/download/install.ps1
DMTools is orchestration layer for enterprise dark factories. It is designed for self-hosted and enterprise environments where teams need repeatable AI-assisted workflows instead of one-off scripts or server-first demos.
DMTools helps platform, delivery, QA, and engineering teams orchestrate end-to-end work such as:
- running MCP tools against Jira, Azure DevOps, GitHub, Confluence, Figma, Teams, SharePoint, and AI providers
- executing reusable jobs and agents for analysis, test generation, reporting, and teammate workflows
- wiring GitHub Actions and other CI/CD pipelines into delivery automation
- packaging focused AI assistant skills for project-level usage in Cursor, Claude, and Codex
| Usage path | What you do with it | Start here |
|---|---|---|
| CLI + MCP tools | Execute direct tool calls and integration operations from the terminal | MCP tools reference |
| Jobs + agents | Run orchestrated workflows such as Teammate, reporting, and test generation | Jobs reference |
| GitHub workflow automation | Run DMTools in CI/CD for ticket processing and teammate flows | GitHub Actions teammate workflow |
| AI assistant skills | Install project-level DMTools skills for agent-driven usage | Skill guide |
All maintained documentation lives under dmtools-ai-docs/.
| Topic | Link |
|---|---|
| Installation and upgrade | references/installation/README.md |
| Configuration overview | references/configuration/README.md |
| Integration setup guides | references/configuration/integrations/ |
| MCP tools reference | references/mcp-tools/README.md |
| Jobs and workflow orchestration | references/jobs/README.md |
| GitHub workflow guidance | references/workflows/github-actions-teammate.md |
| GitHub repository discoverability | references/workflows/github-repository-discoverability-playbook.md |
| Examples | references/examples/ |
| Documentation index | dmtools-ai-docs/README.md |
Latest Version: 
macOS / Linux / Git Bash:
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/install.sh | bashWindows PowerShell:
irm https://github.com/epam/dm.ai/releases/latest/download/install.ps1 | iexJava baseline: Java 17+ for the DMTools CLI and Agent Skill installer flow.
Run the Agent Skill installer from your project root so it can detect project-level skill directories:
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/skill-install.sh | bash
curl -fsSL https://github.com/epam/dm.ai/releases/latest/download/skill-install.sh | bash -s -- --skills jira,githubFor shared base configs and child-agent overrides, use the existing parent config inheritance pattern.
Verify installation:
dmtools --version
dmtools listCreate dmtools.env in your project root (or export the same variables in your shell). You only need the variables for the integrations you actually use — there is no single required set, and the AI provider key is optional (required only for AI-powered jobs and agents).
Jira — full guide
JIRA_BASE_PATH=https://your-company.atlassian.net
JIRA_EMAIL=your-email@company.com
JIRA_API_TOKEN=your-jira-api-tokenAzure DevOps — full guide
ADO_BASE_PATH=https://dev.azure.com/your-organization
ADO_PAT=your-personal-access-token
ADO_PROJECT=YourProjectGitHub — full guide
SOURCE_GITHUB_TOKEN=ghp_your_tokenConfluence
CONFLUENCE_BASE_PATH=https://your-company.atlassian.net/wiki
CONFLUENCE_EMAIL=your-email@company.com
CONFLUENCE_API_TOKEN=your-confluence-api-tokenTestRail — full guide
TESTRAIL_BASE_PATH=https://your-company.testrail.io
TESTRAIL_USERNAME=your-email@company.com
TESTRAIL_API_KEY=your-testrail-api-keyFigma
FIGMA_TOKEN=your-figma-personal-access-tokenAI provider (optional — only needed for AI-powered jobs and agents). Configure one:
# Gemini (free tier available)
GEMINI_API_KEY=your-gemini-api-key
# OpenAI
# OPEN_AI_API_KEY=sk-...
# Enterprise gateway / AWS Bedrock / local models
# DIAL_API_KEY=... # or AWS_ACCESS_KEY_ID + AWS_SECRET_ACCESS_KEY, or OLLAMA_BASE_URLExample combined dmtools.env (Jira + Gemini):
cat > dmtools.env << EOF
JIRA_BASE_PATH=https://your-company.atlassian.net
JIRA_EMAIL=your-email@company.com
JIRA_API_TOKEN=your-jira-api-token
GEMINI_API_KEY=your-gemini-api-key
EOF
chmod 600 dmtools.envFor the full variable list, auth setup (token generation, PAT scopes, base64 credentials), and per-integration options, use the maintained docs instead of copying old snippets from the repository root:
dmtools list
dmtools jira_get_ticket YOUR-123
dmtools run agents/story_development.jsonKeep the latest-release installer flow introduced in DMC-858 when refreshing existing environments.
- Back up
~/.dmtoolsbefore changing anything so you can restore the current wrapper, JAR, and local state if migration fails. - Re-run the installer from
https://github.com/epam/dm.ai/releases/latest/download/install.sh(orinstall.bat/install.ps1on Windows) to replace any legacy bootstrap URLs. - Preserve or merge your existing
dmtools.env,dmtools-local.env, and local configuration instead of overwriting them. - Remove stale aliases, wrapper scripts, and PATH entries that resolve outside
~/.dmtools/bin. - Verify the migrated install with
dmtools --versionanddmtools list. - If anything fails, roll back by restoring the backup copy of
~/.dmtoolsand reactivating the previous wrapper or PATH configuration.
Tagged reinstall example for rollback or migration validation:
macOS / Linux / Git Bash:
curl -fsSL https://github.com/epam/dm.ai/releases/download/v1.7.179/install.sh | bash -s -- v1.7.179Windows:
set DMTOOLS_VERSION=v1.7.179 && curl -fsSL https://github.com/epam/dm.ai/releases/download/v1.7.179/install.bat -o "%TEMP%\dmtools-install.bat" && "%TEMP%\dmtools-install.bat"./gradlew :dmtools-core:shadowJar
./buildInstallLocal.shSee the installation guide for local development installation details.