brier

name

brier

description

Use when the user wants advice or a decision recommendation rather than direct implementation, especially for prompts like "should I", "should we", "which is better", "is it worth it", or "what would you do" about architecture, product, hiring, strategy, or career choices. Prefer the local brier MCP server when available and structure the answer around KPI, option expansion, reference class, disconfirming evidence, numeric forecasts, and a review date.

Brier

Use this skill to turn vague decisions into forecastable choices.

Prefer the local brier MCP server when it is connected.

Workflow

1. If the decision is not stored yet, call create_decision.
2. Use the analyze_decision or review_decision MCP prompts for the canonical structure.
3. Make the analysis explicit:
   • KPI definition
   • KPI resolution metadata
   • option expansion
   • reference class / base rate
   • mechanism or decomposition
   • disconfirming evidence
   • numeric forecasts with 80% confidence intervals
   • a compact forecast summary table
   • review date
4. Persist the result with save_analysis.
   • kpis must be a list of objects shaped like: {"name": "...", "description": "...", "unit": "%|$|days|...", "target": 80, "weight": 1.0, "outcome_type": "binary|count|continuous|percent|currency|score", "resolution_date": "2026-06-30", "resolution_rule": "...", "data_source": "..."}
   • options must be a list of objects shaped like: {"name": "...", "description": "...", "forecasts": [{"kpi_name": "...", "point_estimate": 75, "ci_low": 60, "ci_high": 86, "confidence_level": 0.8, "reasoning": "...", "assumptions": ["..."], "base_rate": 68, "base_rate_source": "...", "inside_view_adjustment": "..."}]}
   • Do not pass KPI or option names as bare strings.
5. If outcomes are known, call score_decision.
6. If the user wants to externalize a forecast into a prediction market, draft it first:
   • Use brier forecast-draft <decision-id> --output forecast-pack.json for stored decisions.
   • Use brier forecast-draft "<forecast question>" --initial-prob <1-99> --resolution-date YYYY-MM-DD --output forecast-pack.json for standalone policy questions.
   • Treat forecast drafts as review artifacts only; do not publish questions or place bets unless the user explicitly asks.

Working Rules

• Do not give a vague recommendation before the KPIs and forecasts are explicit.
• In the first answer, show the forecast results without waiting to be asked again. Include a compact table or bullet summary of each option on the decisive KPIs.
• After the forecast summary, explain how those forecast differences drive the recommendation or prioritization.
• Prefer KPIs that can actually resolve later; avoid labels like fit or quality unless the resolution rule makes them numeric and scoreable.
• Keep the option set broader than the user's initial framing.
• Treat disconfirming evidence as first-class, not cleanup.
• Use the MCP server for persistence instead of leaving the analysis only in chat.

Setup

If the brier MCP server is not connected, add it with:

brier setup claude