Energy research software plays a critical role in research on decarbonization, energy market design, and infrastructure planning. The importance has been reflected by an increased number of initiatives such as FAIR4RS (FAIR for Research Software) principles , which aims to make software FAIR[1]. These developments resonate with the goals of the NFDI4Energy consortium[2], which aims to strengthen research software as a key digital asset in the energy domain. However, researchers often struggle to discover, evaluate, and reuse existing software due to a lack of structured and standardized metadata. In order to understand the current challenges for users, we conducted a workshop titled "Refining Metadata for Energy Simulation Software," held at the 2nd NFDI4Energy Conference[3] in March 2025 in Karlsruhe, Germany. The workshop participants included researchers (both users and developers) from various institutions to explore the role of metadata in improving the findability, accessibility, interoperability, and reusability of energy modeling tools. Participants engaged with real-world use cases that illustrated the difficulty of finding suitable software. In small groups of 4-5 persons each, the participants had to search for software to aid them with their assigned use case using one of the following platforms: Open Energy Platform , SwMath , or GitHub. Each of the groups also had one of the organizers assist them, and collect information. This part helped us highlight the variability of metadata quality and vocabulary usage across platforms, reinforcing the need for a harmonized metadata schema. In the second part of the workshop, the participants were asked to collaboratively fill information about the software they found in the first part in a metadata template from both "a user's" and "a developer's" perspective. The provided metadata template was based on a metadata schema for energy research software which was developed based on . The exercise revealed common ambiguities and inconsistencies in fields like input/output specification and modeling capabilities. Later on, the concept of controlled vocabulary was explained and discussed. Participants then revised their metadata entries by applying controlled vocabulary, promoting clarity and interoperability. The outcomes of the workshop are twofold. First, we collected information regarding how stakeholders look for research software and their practices. Second, we identified fields and terminologies that require harmonization across metadata schemas used in energy informatics. This contribution will outline the design of the workshop, the materials developed (e.g., forms, use cases, slides), and participant feedback. We reflect on how collaborative, use-case-driven approaches can foster metadata literacy and improve software metadata quality in research communities. Furthermore, we discuss implications for registry providers and NFDI consortia aiming to support FAIR research software through standardization . The workshop methodology, materials, and results are openly available and adaptable for other domains facing similar metadata challenges. [1] FAIR stands for Findable, Accessible, Interoperable, and Reusable. [2] https://nfdi4energy.uol.de/sites/about_us/ [3] https://nfdi4energy.uol.de/sites/conference/2025/