Mood shapes how individuals perceive and respond to their environment and influences cognitive processes such as learning, decision-making, and memory. Research on mood-cognition interactions, however, is constrained by mood induction paradigms that produce only short-lived affective changes and lack ecological validity. Given the increasing influence of social media on everyday emotional experience, we developed a novel social media-based mood induction paradigm resembling real-world digital environments. The primary aim was to assess whether experimentally designed positive and negative social media feeds induce selective changes in mood. To maximize experimental control, the feeds were carefully matched in image categories and content structure while differing systematically in emotional valence. In addition, we explored whether a brief mindfulness breathing intervention moderates the mood induction effects and whether induced affective states influence reward-based learning and decision-making. Healthy participants were randomly assigned to either a brief mindfulness or visually matched control intervention. Each participant completed both positive and negative social media feeds, each followed by a reward-based decision-making task. Valence and arousal ratings were collected repeatedly throughout the experiment. Preliminary pilot data (N = 15) demonstrate the feasibility of the paradigm and show that the social media feeds reliably induced mood changes in the intended direction. Participants reported significantly increased mood following the positive feed and significantly reduced mood following the negative feed. Importantly, effects were specific to affective valence, whereas arousal remained unchanged, indicating selective manipulation of emotional valence rather than general emotional activation. Notably, mood changes persisted over time and appeared longer-lasting than those observed with induction procedures previously used in our lab. The mindfulness intervention did not significantly moderate the mood induction, and no significant downstream effects on reward learning performance were observed. This pilot study establishes a novel, ecologically valid social media-based mood induction paradigm and provides a promising framework for future studies investigating how affective states influence cognition and how such interactions may inform research on mental health.