reversa-arbiter

Scores ideation options against risks and records the user's decision in decision.md.

Updated May 22, 2026
One-click install
npx skills add https://github.com/GleisonOliveira/dockerpilot-mcp --skill reversa-arbiter-gleisonoliveira
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: reversa-arbiter
Source: https://github.com/GleisonOliveira/dockerpilot-mcp/tree/main/.agents/skills/reversa-arbiter
Command: npx skills add https://github.com/GleisonOliveira/dockerpilot-mcp --skill reversa-arbiter-gleisonoliveira

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams often stall after brainstorming because no one converges the options into a documented, human-owned decision. This Skill scores each option against fixed criteria, produces an explicit recommendation with trade-offs, and records the user's choice as a structured decision document. ## Core Features & Use Cases - Criteria-Based Scoring: Scores each option from options.md on job-to-be-done fit, effort, residual risk, and legacy cost using fixed 1-5 scales. - Explicit Trade-Off Recommendation: Recommends one option while naming what is lost, under what condition the recommendation changes, and what to test before committing. - Human Decision Recording: Presents a decision menu, waits for explicit user input, and writes the outcome to decision.md, registering any divergence between recommendation and choice. - Use Case: After running brainstorm and risk-analysis steps in an ideation session, invoke this Skill to converge on a single option and produce an auditable decision.md before moving to specification. ## Quick Start Ask the agent to run the reversa arbiter step to score the options in the active ideation session and record my decision in decision.md.

Frequently Asked Questions about reversa-arbiter

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I decide between multiple product options after brainstorming?▼

Score each option on job-to-be-done fit, effort, residual risk, and legacy cost using fixed 1-5 scales, then sum without weighting. The Skill recommends the highest scorer with explicit trade-offs, but the final decision is always recorded as a human choice.

What files does the reversa arbiter need before it can run?▼

It requires an active ideation session with framing.md, options.md, and risks.md present in the session directory. If risks.md is missing, it stops and instructs you to run the challenger step first.

What happens if I choose an option different from the recommendation?▼

Your choice is recorded as the decision and the recommendation is logged as a divergence in decision.md. The scoreboard is never adjusted to match your choice, since the divergence itself is valuable information.

Can the arbiter write decision.md without my explicit answer?▼

No. It presents a decision menu and waits for your response before writing anything. Silence is never treated as acceptance, and if decision.md already exists it asks for explicit confirmation before overwriting.

What happens when two options tie in the scoring?▼

A tie is treated as a legitimate result. The Skill records the tie and explains what decides it in practice rather than inventing an artificial tiebreaker.