The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions define an ExecutionProxy Interface that uses the rtk prefix for terminal actions (e.g., rtk npm test).
Evidence: This is presented as a mandatory protocol in SKILL.md to ensure deterministic execution and track computational overhead within the agent's environment.
[SAFE]: The skill includes explicit security guidelines under the "Security & Multi-Agent Hygiene" section.
Evidence: Instructions mandate that web content and external data be treated as hostile and require the redaction of secrets and PII (Personally Identifiable Information) before sharing context with subagents.
[SAFE]: External references to A/B testing tools and calculators target well-known, reputable industry services.
Evidence: Links provided in references/sample-size-guide.md include Evan Miller's calculator, Optimizely, VWO, and AB Test Guide, all of which are standard tools for statistical analysis.
[SAFE]: The skill enforces a strict methodology for experiment design, including hypothesis locking, mandatory sample size estimation, and predefined success metrics, which promotes technical integrity and data reliability.

ab-test-setup