{"skill":{"slug":"deep-research-prime","displayName":"Deep Research","summary":"AI agent for deep research tasks","tags":{"latest":"0.1.0"},"stats":{"comments":0,"downloads":1756,"installsAllTime":10,"installsCurrent":9,"stars":2,"versions":1},"createdAt":1771141439145,"updatedAt":1777525159392},"latestVersion":{"version":"0.1.0","createdAt":1771141439145,"changelog":"deep-research-prime 0.1.0\n\n- Initial release of the skill specializing in deep research task decomposition and orchestration.\n- Implements strict division of labor: the leader agent plans, delegates, and synthesizes—never directly gathers information.\n- Enforces user input as the sole source of truth; never alters terms provided by users.\n- Mandates parallel task decomposition (2–6 orthogonal subtasks per project) with one-round research execution.\n- Details specific subordinate agents for internet, academic, financial, and social media information gathering.\n- Establishes robust principles for task verification, expansion, and comprehensive information evaluation.","license":null},"metadata":null,"owner":{"handle":"realroc","userId":"publishers:realroc","displayName":"YuPeng Wu","image":"https://avatars.githubusercontent.com/u/44936809?v=4"},"moderation":null}