Contemporary agile development processes optimize for velocity and functional completeness, yet systematically disregard ethical considerations. This dilemma exposes a significant ethics gap, as organizations lack structured mechanisms to embed ethical deliberation within agile practices without impeding agility. Anchored by the ISO/IEC/IEEE 24748-7000:2022 standard on Value-Based Engineering, we advance the integration of Value-Based Engineering throughout the agile requirements life cycle by introducing the Ethics Filter Framework (EFF), a bridging intervention designed to translate ethically indifferent or ambiguous User Stories into ethically coherent requirements through explicit value elicitation and stakeholder reflection.  This urgency for this intervention is given by the rise of autonomous AI coding agents. Autonomous AI coding agents dramatically accelerate software production by transforming natural-language utterances directly into executable code, which means that any missing ethical requirement is instantly propagated into system logic without human oversight. In this setup, the User Story is the final actionable moment where ethical requirements can still be concretely integrated.  Implementing Value-Based Engineering principles in User Stories, therefore, is a failsafe to prevent biased and unsafe behaviors, whether by means of Human or Artificial Intelligence, from being silently operationalized into system architectures. Our methodology reconceptualizes ethical evaluation as integral to the definition of quality in software requirements, which are formulated as User Stories in Agile practice.  By establishing Values as a core component of User Stories, the Ethics Filter Framework (EFF) provides a structured approach to ensure that applications not only demonstrate functional and syntactic adequacy, but also uphold human dignity, autonomy, and fairness, thus advancing the notion of making ethics agile.