Ideas on governing AI without slowing teams down.
Perspectives from the Tracelet team on visibility, control and safe AI adoption across security, engineering and leadership.
- One line, every email: the postmark-mcp backdoor
How a copycat MCP server quietly BCC'd its users' mail to a stranger, why nobody saw it, and what Tracelet can and can't do about the next one.
Tracelet team · 9 min read - Introducing Tracelet: visibility and control for AI coding tools
AI coding tools are already on your developers' machines. Tracelet shows what they do, applies the rules you set, and gives security, engineering and leadership one shared view.
Tracelet team · 2 min read - Shadow AI is a visibility problem, not a discipline problem
Developers reach for unsanctioned AI tools because they work. Policies that ignore that drive usage out of sight. Start by seeing what is already happening.
Tracelet team · 2 min read - AI governance that stays in your environment
Governing AI tools means recording sensitive decisions about code, credentials and data. Those records belong inside your own environment, captured with restraint.
Tracelet team · 1 min read
All posts
- One line, every email: the postmark-mcp backdoorTracelet team · 9 min read
- Introducing Tracelet: visibility and control for AI coding toolsTracelet team · 2 min read
- Shadow AI is a visibility problem, not a discipline problemTracelet team · 2 min read
- AI governance that stays in your environmentTracelet team · 1 min read
- What leadership should measure about AI adoptionTracelet team · 1 min read
- Guided, not blocked: why AI governance should speed developers upTracelet team · 1 min read