/mass-line

Gathers multi-source intelligence via grassroots survey, synthesizes consensus/conflicts/gaps, and returns validated findings with recommendations for next-round feedback.

Generates a multi-perspective analysis of a topic by scanning the vault for context, building 3-5 stakeholder perspectives with concerns and tensions, and synthesizing agreements and conflicts.

Analyzes a question, plan, or idea from multiple cognitive perspectives in parallel, producing a synthesized Council Report. Supports council size and deep mode flags.

Initiates isolated voting among discuss session participants: sends the same question privately to each, collects structured responses via LLM extraction, and generates a Markdown report grouped by stance. Also supports list, report, collect, next-round, and cancel subcommands.

Assembles AI agent council with dynamic personas to generate parallel opinions on a query, anonymously rank them, and synthesize weighted final response.

Runs an arbiter-mediated consensus loop where GPT, Gemini, Grok, and optional OpenRouter delegates review a plan while Claude commits a blind verdict, adjudicates issues, and revises until cross-model agreement is reached.