What we shipped at Heym in the last 30 days

@Heym update 👋

The last 30 days have been some of our busiest yet. @mbakgun

We shipped a broad set of improvements across security, observability, workflow migration, dashboards, agent capabilities, and integrations.

A few highlights:

🔒 Security hardening
OS-level sandboxing for agent code, OAuth hardening, SSRF-protected downloads, parameterized queries, stricter resource access, and private vulnerability reporting.

🔎 Better observability
OpenTelemetry tracing, execution timelines, token usage, and model cost visibility.

🔄 Easier workflow migration
AI Convert can bring existing automations, including n8n workflows, into Heym without rebuilding everything manually.

📊 Built-in operational tooling
Dashboards, DataTables, runbooks, and agent skills that can run Python and generate real files and reports.

🔌 New integrations
Amazon S3, Supabase, Discord, ElevenLabs, and ChatGPT through MCP.

We are building Heym as a self-hosted AI workflow platform that teams can inspect, secure, and operate in production.

We would really value feedback from the Product Hunt community:

Which of these areas matters most to you right now?

Security
Observability
Migration
Integrations
Agent capabilities

Product: https://heym.run
GitHub: https://github.com/heymrun/heym

Which area should we prioritize next: security, observability, migration, integrations, or agent capabilities?