roast
01Command
A questioning mode that stops giving ordinary answers and keeps asking deeper questions about the user's thinking.
24 tagged critical thinking, measured the same way as everything else here.
Browse within: Reasoning 8document-analysis 8analysis 6aquinas 6claude-code-skills 6
Command
A questioning mode that stops giving ordinary answers and keeps asking deeper questions about the user's thinking.
Command Claude Code
Command "guru-pk" from MitsudoAI/guru-pk-mcp, covering guru-pk expert debate system, 语言模板定义, 中文模板 (zh), 英文模板 (en) and 日语模板 (ja).
Command
Generate or view the AITS 2.0 decision dashboard. Shows the current state of an ongoing analysis (agents completed, findings, conflicts, open gates) or produces a self-contained HTML dashboard for a completed analysis. The dashboard includes agent flow graph, risk heatmap, confidence meter, conflict timeline…
Command
Divergent AITS 2.0 brainstorming in review mode. Activates the Meta-Orchestrator with a creative-first sequence (Creative-Generative → Emotional-Intuitive → Foresight → Synthesis) and runs the full analysis without voluntary checkpoints. At the end, presents everything with a drill-down review interface. Default for…
Command
Full AITS 2.0 analysis in supervised mode. Activates the Meta-Orchestrator with checkpoint after every agent. The user can approve, correct, redirect, redo, or switch to autonomous at each step. Default for high-stakes, irreversible decisions. Consults pattern library and memory, matches playbooks, enforces all seven…
Command
Decompose through Cartesian Reductionism (methodical doubt).
Command
Classify decisions as one-way or two-way doors and determine required analysis depth.
Command
Multi-framework epistemic analysis — compare, suggest, or pick frameworks.
Command
Challenge the hidden assumptions in a document.
Command
Examine the evidence and sources behind the claims.
Command
Run all six lenses in parallel and synthesize the findings.
gagharutyunyan1993/methodology-toolkit
Command
Force methodology-driven analysis of a task, optionally with a specific method.
Command
Cancel an active Anvil debate.
Command
Show current Anvil debate status.
Command
Start an adversarial debate to stress-test an idea.