Scale guard clause complexity to false-positive cost — low-stakes triggers need zero guards, high-consequence ones need multiple

Why This Is a Rule

Not all false positives cost the same. A meditation trigger firing at the wrong time costs you 30 seconds of noticing and dismissing it — negligible. A "send the escalation email" trigger firing in the wrong context could damage a professional relationship — significant. The guard clause investment should match the cost of getting it wrong, not follow a uniform standard.

This is the behavioral version of the engineering principle that safety-critical systems need more validation layers than convenience systems. A light switch doesn't need a confirmation dialog; a nuclear launch button does. Similarly, a "drink water" trigger needs zero guards (worst case: you drink water when you weren't thirsty), while a "confront the performance issue" trigger needs context checks (worst case: you initiate a difficult conversation at an inappropriate moment).

The matching principle prevents two complementary errors: over-guarding low-cost triggers (adding deliberation overhead that makes the agent slower than not having it) and under-guarding high-cost triggers (automating consequential behavior without adequate context verification).

When This Fires

• When deciding how many guard clauses a trigger needs (complements When false positives exceed 30%, add guard clauses — context checks that must pass before the action executes)
• When reviewing existing agents and finding that all triggers have the same number of guards regardless of consequence
• When a low-cost trigger has become bogged down with unnecessary checks
• When a high-consequence trigger fires inappropriately and causes damage

Common Failure Mode

Applying uniform guard complexity across all agents: either zero guards for everything (efficient but reckless for high-stakes agents) or multiple guards for everything (safe but creates unnecessary friction for low-stakes agents). The optimal guard count varies by agent — match investment to risk.

The Protocol

(1) For each trigger, assess the cost of a false positive: what happens if the agent fires when it shouldn't? (2) Low cost (drink water, check posture, take a breath) → zero to one guard clause. Let the trigger fire freely; false positives are harmless. (3) Medium cost (start a conversation, switch tasks, send a message) → one guard clause. A quick context check before action. (4) High cost (send difficult email, make irreversible commitment, confront a conflict) → two to three guard clauses. Multiple context checks ensuring the timing, audience, and situation are right. (5) Never add more guards than the cost justifies. The deliberation overhead of each guard clause is a real cost — it must be less than the expected cost of the false positives it prevents.