AGENTFORCE-9.2: Cross-Topic Instruction Boundary

🔴 High · Instruction Integrity

Detects when a topic's instruction text references another topic's name or attempts to override the planner's topic-selection logic. Cross-topic instruction bleeding allows adversarial metadata to manipulate conversation routing, effectively hijacking agent flow without modifying the transition graph.

Details

Field	Value
Rule ID	AGENTFORCE-9.2
Severity	High
Category	Instruction Integrity
Compliance	EU_AI_ACT_HIGH_RISK, NIST_AI_RMF

Remediation

Refer to the SquireX documentation for remediation guidance specific to AGENTFORCE-9.2.

See Also

• Rule catalog index
• SquireX scan documentation