← ClaudeAtlas

hivemind-architect-strategistlisted

Core skill for the hiveminder agent. Provides scanning, analysis, and synthesis capabilities for architecture review and strategic planning. Research-fronted and evidence-oriented approaches for high-quality decision making.
shynlee04/hivemind-plugin · ★ 51 · Code & Development · score 65
Install: claude install-skill shynlee04/hivemind-plugin
# HiveMind Architect Strategist **Core principle:** Research-fronted. Evidence-oriented. Clean architecture. Ruthless quality. Never decide without data. Never claim without proof. ## When to Use - Starting architecture review or code audit - Planning strategic refactoring decisions - Evaluating technical debt and migration paths - Synthesizing findings from multiple code scans - Making evidence-based recommendations ## Plugin Dependency This skill is part of the `hivemind-context-governance` plugin. It provides core capabilities for the hiveminder agent. When installed to client projects via `npx hivemind init`, this skill becomes available in: - `.opencode/skills/hivemind-architect-strategist/SKILL.md` (project-level) - Or `~/.config/opencode/skills/hivemind-architect-strategist/SKILL.md` (global) ## Core Philosophy | Principle | Meaning | |-----------|---------| | **Research-fronted** | Always investigate before deciding. Scan code, check docs, validate assumptions. | | **Evidence-oriented** | Claims require proof. "This is broken" → show the error. "This should work" → cite the docs. | | **Clean architecture** | Follow SOLID principles. Respect layers: tools/lib/hooks/schemas. No shortcuts. | | **Ruthless quality** | No compromises on standards. If it doesn't pass tests, it doesn't ship. | ## Scanning Capabilities ### Using HiveMind Custom Tools (requires plugin) ```typescript // Deep hierarchy scan scan_hierarchy({ action: "analyze" }) // Recall past decision