From autodecision
Compare two decisions side-by-side — either run fresh or compare existing runs
npx claudepluginhub harshilmathur/autodecision --plugin autodecision# /autodecision:compare Compare two decisions or decision variants side-by-side. ## Two Modes ### Mode 1: Fresh Comparison (run quick mode on both) Runs `/autodecision:quick` on BOTH options, then produces a side-by-side comparison. If one option is clearly better, recommend running the full loop on it. ### Mode 2: Post-Facto Comparison (compare existing runs) Reads two existing Decision Briefs from `~/.autodecision/runs/` and produces a structural comparison. Works for both quick and full mode runs. ## Execution ### Fresh Comparison 1. Parse the two decision statements from the ...
/decision-tree-explorerExplores decision branches with probability weighting, expected value analysis, and scenario-based optimization.
/optionsCompares options side-by-side against weighted criteria, scores them, and recommends the best fit. Interactively asks questions to clarify decision context if needed.
/decision-quality-analyzerAnalyzes team decision quality via scenario testing, bias detection, and process optimization using specified criteria.
/war-roomConvenes multi-LLM expert panel to pressure-test strategic decisions via reversibility assessment, adversarial review, and deliberation modes, producing synthesized recommendation.
/analyzeAnalyzes ADR options against criteria with Low/Medium/High risk ratings, generates comparison matrix with rationales and mitigations, summarizes trade-offs, and saves state.
Share bugs, ideas, or general feedback.
Compare two decisions or decision variants side-by-side.
/autodecision:compare "Cut pricing 20%" vs "Cut pricing 10%"
/autodecision:compare "Launch in US" vs "Launch in SE Asia"
Runs /autodecision:quick on BOTH options, then produces a side-by-side comparison.
If one option is clearly better, recommend running the full loop on it.
/autodecision:compare pricing-cut-20pct-full vs market-expansion-full
/autodecision:compare --existing slug-a vs slug-b
Reads two existing Decision Briefs from ~/.autodecision/runs/ and produces
a structural comparison. Works for both quick and full mode runs.
DECISION-BRIEF.md files from ~/.autodecision/runs/{slug}/.effects-chains.json files for structural comparison.Write to ~/.autodecision/runs/comparison-{slug-a}-vs-{slug-b}.md:
# Decision Comparison: {Decision A} vs {Decision B}
Generated: {date}
## Summary
| Dimension | {Decision A short name} | {Decision B short name} |
|-----------|------------------------|------------------------|
| Recommendation | {action} | {action} |
| Confidence | {level} | {level} |
| Mode | {quick/full} | {quick/full} |
| Hypotheses explored | {N} | {N} |
| Effects (1st order) | {N} | {N} |
| High-consensus effects | {N} (if full) | {N} (if full) |
| Iterations | {N} | {N} |
| Converged | {yes/no} | {yes/no} |
## Risk Profile Comparison
| Risk Type | {A} | {B} |
|-----------|-----|-----|
| Execution risk | {HIGH/MED/LOW} | {HIGH/MED/LOW} |
| Competitive risk | {HIGH/MED/LOW} | {HIGH/MED/LOW} |
| Financial risk | {HIGH/MED/LOW} | {HIGH/MED/LOW} |
| Reversibility | {HIGH/MED/LOW} | {HIGH/MED/LOW} |
| Time to value | {timeframe} | {timeframe} |
## Shared Effects
Effects that appear in BOTH decisions (same or similar effect_ids):
{list with probability comparison: "acq_increase: A=0.60, B=0.45"}
## Unique to {A}
Effects or insights that only appear in Decision A:
{list}
## Unique to {B}
Effects or insights that only appear in Decision B:
{list}
## Assumption Overlap
Assumptions shared between both decisions:
{list — these are your "organizational assumptions" that recur across decisions}
## Key Tradeoffs
{What do you gain by choosing A over B? What do you lose?}
{What do you gain by choosing B over A? What do you lose?}
## Recommendation
{If one is clearly better: "Decision A is the stronger option because..."}
{If it's a genuine tradeoff: "This depends on [specific factor]. If X, choose A. If Y, choose B."}
{If the decisions are not comparable (apples vs oranges): "These are independent decisions — the comparison surfaces shared assumptions but they don't compete."}
If both decisions reference the same assumptions (e.g., both reference
price_sensitivity_moderate), note it in the comparison. Shared assumptions
across different decisions are candidates for the assumption library — they
represent organizational beliefs worth tracking and validating.