Webinar: what your agents actually execute
Online — 11:00 AM
45 minutes: put the Recorder in observe mode in front of an MCP server, read the evidence stream that comes out, then verify a bundle offline with sealr-verify.
Events
Let's meet
Workshops, webinars, conferences: we come to talk about what AI agents execute in production, the evidence you can produce from it, and the guardrails you can put in front of it.
Calendar
Upcoming
Calendar in progress. The three entries below are layout examples: none of these dates is confirmed and no registration is open. They will be replaced by the real calendar before launch.
Workshop: writing a guard policy
Paris, 8th arr. — 6:30 PM
SQL Guard, Terraform Guard and Git Guard in practice: DELETE with no WHERE, a Terraform plan that destroys a protected resource, a force-push to a protected branch. Write the rules, then simulate them against real records.
FIC 2027 — Lille
Lille Grand Palais
End-to-end demo: interception, verdict, human approval, timestamped checkpoint, then the export of an .seal bundle a third party verifies in front of you.
Media library
Replays
No replay is available yet. The two entries below illustrate the layout and will be replaced by the sessions we actually record.
45 min
Execution evidence: where to start
The available interception surfaces — MCP proxy, agent hooks, CLI wrapper — and what each one covers or lets through.
Browse the library38 min
Going from observe to enforce without blocking your teams
Picking protected resources, tuning approvals, keeping a break-glass path: the sequence that keeps a guardrail from becoming an outage.
Browse the library