Integrations

Integrations

🌍 English narrative is auto-generated on release. See the French version for the current full descriptions.

WorkPilot AI connects to third-party systems for issue tracking, source hosting, and tool execution (MCP).

Integration	Path
azure_devops	apps/backend/integrations/azure_devops
graphiti	apps/backend/integrations/graphiti
linear	apps/backend/integrations/linear
team_bot	apps/backend/integrations/team_bot.py
windsurf_mcp	apps/backend/integrations/windsurf_mcp.py
windsurf_proxy	apps/backend/integrations/windsurf_proxy

Auto-generated from apps/backend/integrations/ · 6 integrations.

Integrations

WorkPilot AI connects to the platforms teams already use for issue tracking, code review, and collaboration, so agents can work directly against real backlogs and pull requests rather than an isolated task list.

Platform	Capabilities
GitHub	Import issues, AI-powered investigation, PR review, batch review wizard, auto-PR creation
GitLab	Issue and merge request management with AI severity categorization
Azure DevOps	Work item import, PR review, batch operations — experimental: several backend endpoints still return mock data
Linear	Bulk issue import with team/project filtering
Jira	Issue management integration
Windsurf	IDE integration via the Connect protocol
Microsoft Teams	Webhook-based notifications for task completion and PR creation

On top of these, the MCP Marketplace lets you browse, install, and configure Model Context Protocol servers, and Custom MCPs support defining and hosting your own MCP servers with local authentication — extending what agents can reach without waiting on a built-in connector.

Semantic code search is provided by the bundled grepai integration: natural-language search over the codebase, vector-embedding matches, and call-graph tracing via grepai trace, with automatic fallback to standard grep when the grepai server isn't running.

On the AI side, WorkPilot AI is provider-flexible rather than locked to a single vendor. It supports Anthropic Claude (OAuth or API key), OpenAI, Google Gemini, Grok/xAI, Ollama (local), Azure OpenAI, GitHub Copilot, and any OpenAI-compatible custom endpoint (e.g., z.ai for GLM models). Multi-account switching lets you register several profiles per provider, with WorkPilot AI automatically failing over to an available account when one hits a rate limit — keeping autonomous builds running without manual intervention.

Together, these integrations let the multi-agent pipeline plan, implement, and validate work using the same issue trackers, repositories, and AI backends already in place, rather than requiring teams to adopt a separate system of record.

---

➡️ Continue: AI providers