Launch course / Module 00 / Free as released
How agents fail: Module 00.
Planned0 of 1 modules accessible. Joining Skool does not unlock planned modules.
First modules in this course December 14, 2026. This module's release date will be confirmed before it opens.
Free, shared with every track. An agent combines a model, an execution loop, tools, and memory. Each introduces a trust boundary. The case file references Microsoft's CVE-2026-24307 record for M365 Copilot. You leave with an untrusted-input map for one agent.
Module topics
- Agent components and trust boundaries
- Untrusted-input map
- Agent register with owner, actions, resources, credentials, and approval level
Related risks
- ASI01 Agent Goal Hijack
Lab and deliverable
Untrusted-input map and agent register.
Study time: 2 to 4 hours, plus practice. Each module confirms lab requirements and any cloud or API costs before it opens.
Signed module record
The named practitioner author, AI-assistance disclosure, model and version, hand verification and signed sign-off will be published with the release.
This page is the public module home, not an accessible lesson or a signed release record.
Return to Module 00 in the course / Pathways / Release schedule / Provenance policy