LunaPrompts is a browser-based technical screening platform designed specifically for hiring AI engineers. The service addresses the challenge of evaluating candidates' practical AI skills by offering AI-native assessments rather than traditional algorithm puzzles. It targets hiring leaders and teams seeking to assess abilities in areas such as prompt engineering, agent code, retrieval-augmented generation (RAG), SQL, and system design, reflecting the contemporary demands of AI-focused roles.
The platform provides a range of assessment types, each with its own grading pipeline. 5-flash, with full control over hyperparameters such as temperature, top_p, max_tokens, and frequency and presence penalties. Agent code challenges allow for multi-file Python projects within a Judge0 sandbox, supporting tool-call loops and environment files. Algorithm problems are available in nine programming languages with both public and hidden tests, while SQL assessments use in-memory DuckDB on real datasets. System design is tested through a visual canvas graded by rubric, and multiple choice questions serve as a quick initial filter.
LunaPrompts emphasizes that AI is integral to the assessment process; candidates interact directly with language models, and their ability to prompt, iterate, select models, and debug is evaluated. The grading focuses on the candidate's craft and judgment rather than merely the final answer. Anti-cheating measures are in place to prevent impersonation and outsourcing, including in-browser MediaPipe FaceMesh face detection, fullscreen enforcement, tab-switch detection, clipboard and developer tools blocking, and auto-submission based on configurable thresholds.
The platform operates entirely through the browser and does not require integration with applicant tracking systems to begin. Candidate invites can be sent via CSV or XLSX files. Greenhouse, Lever, Ashby integrations, and SAML SSO are planned for future releases. Pricing begins with a pilot phase for a single role and 10 to 30 candidates, with no commitment required. Ongoing annual plans are seat-based for assessment teams and offer volume tiers based on the number of candidates evaluated.
LunaPrompts sits in PulseGate's HR & recruiting category. It focuses on assessing the real-world AI engineering skills of candidates beyond traditional coding tests. It is built as a B2B product for tech recruiters and hiring managers for AI engineering roles. LunaPrompts is a paid product. LunaPrompts is available on the web.
LunaPrompts builds and maintains LunaPrompts, and the product first shipped in 2026. Development happens publicly on GitHub with 55 stars and 430 commits in the last 90 days. Key capabilities include prompt engineering tests, agent code assessments, and SQL and system design challenges.
Latest indexed changes and source events
luna-prompts/skillnote verified by the PulseGate indexer
Other apps tracked under the same category.
