SAFE-MCP: Industry Specific Agentic Workflow Hackathon

​AI agents are being deployed across healthcare, fintech, manufacturing, biotech, and infrastructure systems introducing new attack vectors that existing security frameworks don’t fully cover.

​This is a weekly online hackathon-style working session where engineers pick an industry-specific agentic workflow use case and write it end-to-end: document the workflow, map it to real AI agent attack techniques (SAFE-T), and propose practical mitigations. The output is published as real GitHub contributions in SAFE-MCP/safe-agentic-use-cases (issues + PRs).

​Call to Action:
If you build, secure, or operate AI agents in production and want to help author industry guidance on how these systems are attacked and defended join the weekly session and contribute a use case.

​Format:

• ​One weekly online session

• ​Work on one use case per participant (participants can also collaborate on the same use case)

• ​Write: agentic workflow → SAFE-T mapping → mitigations

• ​Contribute directly in GitHub (issue / pull request) as the written output

​Participants may contribute on behalf of their company or in a personal capacity.