Blog
Guides, product updates and engineering notes on building AI agent workflows you can watch run.
- What is an AI agent workflow? A practical guideAn AI agent workflow puts AI agents inside a graph of ordinary steps: triggers, tools, branches and outputs. What it is, when to use one, how to build it.
- How to give a Claude agent tools: HTTP, apps and MCPA hands-on guide to Claude tool calling in Worfilo: attach HTTP, app, API or MCP tools to an agent, bound the tool loop, and branch on structured output.
- Build AI workflows from Claude Code, Cursor or VS CodeThe Worfilo MCP server lets Claude Code, Cursor and VS Code plan, build, test and publish AI workflows, then call them from your code with a scoped key.
- Mask personal data before it reaches the modelWorfilo's new Mask PII node strips names, emails, cards and national IDs from text before a Claude agent reads it. How it detects PII and how we test it.
- Run AI agent workflows on a scheduleWorfilo triggers run a workflow's published version on a cron schedule in your timezone: a daily digest, an hourly sync, a monthly report. How they work.
- MCP servers, with a marketplace to find themWorfilo now connects to MCP servers: browse a public marketplace or add one by URL, then call its tools from a node or hand the server to an agent.
- Connectors, custom APIs and AI workflow generationWorfilo now connects to GitHub, Slack, Notion, Linear and Tavily, calls any API with any auth, and drafts whole workflows from a description.
- Introducing a visual editor for AI agent workflowsWhy I built Worfilo, a visual editor for AI agent workflows you can watch run: what it does today and what comes next, from developer Oshadha Vimukthi.
- Designing a workflow engine you can watch runHow Worfilo's executor works: one rule for active and dead edges, run events streamed over SSE, and graph snapshots that make run replay trustworthy.