CSE Roadmap

Common tier · about 3 hours · 12 parts

Part 6 of 12 · 2 min read

What the example found

The STRIDE threat count by category, the flagged crossings before and after controls, and the mitigation search results.

How to read the STRIDE-count table: each row is one threat category from the mechanical per-element enumeration against the fictional support agent's 7 nodes and 6 flows.

Category Count
Spoofing 5
Tampering 10
Repudiation 7
Information disclosure 10
Denial of service 10
Elevation of privilege 2
Prompt injection (added) 2
Total 46

What to notice: the two added prompt-injection threats are a small fraction of the 46 total. But they are also the two threats that a vanilla STRIDE checklist could never have produced on its own. Count and importance are not the same axis.

How to read the crossing table: before and after the same three controls (validation, quarantine, authorisation plus approval) are added to the flows the structural rule flagged.

Step Result
Flows into higher trust with no control, before customer to agent, web to agent, agent to tools
After validation, quarantine, authorisation and approval none

What to notice: three named flows going to zero is a complete result, not a partial improvement. The structural rule that found them is the same rule used to confirm they're gone, so there's no separate "did it actually work" question left open.

How to read the mitigation-search table: each row applies a different set of controls to the four-path attack tree and reports the resulting probability the goal still succeeds.

Mitigation set Probability the goal succeeds
None 43.3%
Egress allowlist only 33.2%
Least privilege only 28.9%
Quarantine only 33.2%
All seven 0%
Cheapest blocking set (least privilege, tenant filter, trace ACL; cost 5 of 12) 0%

What to notice: the cheapest blocking set reaches the exact same 0% as applying all seven controls, at under half the total possible cost. The search found that most of the seven available controls were redundant once the right three were chosen together.

{{svg:dfd|Seven nodes and six flows with trust zones; flows without controls toward higher trust are red.}} {{svg:results|Probability by attack path, mitigation what-ifs and the cheapest blocking set.}}