Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Multi-Criteria Scoring Campaign — evaluate and rank candidates against multiple weighted criteria using AHP, BWM, TOPSIS, VIKOR, ELECTRE, PROMETHEE, MAUT methods.
.claude/skills/yogsoth-ai-convergence-multi-criteria-scoring/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 161% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 547% | 0% |
| case-06 | ✓→✗ | ▼ Worse | 199% | 0% |
| case-09 | ✓→✗ | ▼ Worse | 40% | 0% |
| case-10 | ✓→✗ | ▼ Worse | -41% | 0% |
Evaluate and rank candidate alternatives using multi-criteria weighted assessment. Supports mainstream MCDA methods including AHP, BWM, TOPSIS, VIKOR, ELECTRE, PROMETHEE, and MAUT, covering the full workflow from criteria definition, weighting, scoring, to sensitivity analysis.
| Signal | Strategy | |--------|----------| | "What criteria? How to score?" / select best option / rank by criteria | best-option-selection | | "Full ranking" / full ranking / league table / priority list | full-ranking | | "Categorize" / categorize / A/B/C grading / compliance sorting | category-sorting | | "Eliminate non-qualifying" / go/no-go / safety screening / veto | non-compensatory-screening | | "Determine weights" / weight negotiation / criteria importance | weight-elicitation |
| Strategy | Purpose | Budget | |----------|---------|--------| | best-option-selection | Select best option from candidates | M | | full-ranking | Produce complete priority ranking | M | | category-sorting | Classify candidates into predefined categories | M | | non-compensatory-screening | Eliminate non-qualifying candidates | S | | weight-elicitation | Determine criteria weights | S |
| Tactic | Purpose | |--------|---------| | scoring-matrix-construction | Define criteria → assign weights → score → aggregate → sensitivity check | | screening-then-scoring | Non-compensatory elimination first → then fine-grained scoring of survivors | | multi-method-triangulation | Multi-method comparison → identify method-sensitive options |
| SOP | Purpose | |-----|---------| | criterion-definition | Extract evaluation criteria from research goals and candidates | | weight-elicitation-sop | Compute criteria weights using specified method | | alternative-scoring | Score candidates against each criterion | | normalization | Normalize the score matrix | | dominance-check | Identify dominated and non-dominated alternatives | | threshold-setting | Set minimum thresholds for each criterion | | conjunctive-filter | Eliminate candidates failing thresholds | | rank-comparison | Compare consistency of multiple ranking results | | method-sensitivity-report | Analyze impact of method choice on rankings | | scoring-synthesis | Synthesize scores, rankings, and sensitivity into final recommendation |
| Metric | Target | |--------|--------| | Criteria count | 5-8 | | Candidate count | 8-15 | | Weight method comparison | >=2 methods | | Sensitivity analysis | >=3 parameter perturbations |
| Tool | Usage | |------|-------| | vault_search | Retrieve existing evaluation criteria and historical scores | | vault_query_graph | Query relationships and dependencies between alternatives | | vault_add_edge | Record evaluation result relationships |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Strategy | When to use | | --- | --- | | best-option-selection | Select the single best candidate from a set using WSM, TOPSIS, AHP, MAUT, or VIKOR methods. | | category-sorting | Classify candidates into predefined categories using ELECTRE-Tri, FlowSort, AHPSort, or DRSA methods. | | full-ranking | Produce a complete ordering of all candidates using PROMETHEE I/II, ELECTRE III, or MAVT methods. | | non-compensatory-screening | Eliminate non-qualifying candidates using conjunctive rules, dominance filtering, lexicographic ordering, or veto thresholds. | | weight-elicitation | Determine criteria weights using AHP, Swing, BWM, MACBETH, or Simos methods. |
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | context-checkpoint | Append research process and results to the current Phase's context file. Covers both process and results with genuine substance. Use this skill at plan-designated checkpoint points — typically after each strategy completes or at key decision nodes within a research Phase. | | context-init | Create a new context file for a research Phase. Called once at Phase start to initialize the file that subsequent context-checkpoint calls will append to. Use this skill whenever a new research Phase begins and a fresh context file is needed. | | convergence-paper-overview | Paper landscape scan at abstract level — discover MCDA, voting theory, Delphi, and optimization methodology papers. | | convergence-paper-research | Full-text deep reading of methodology papers — complete understanding of algorithms, proofs, and implementation details. | | convergence-paper-search | Paper AI summary reading — deeper understanding of specific methodology papers without full-text commitment. | | convergence-saturation-detection | Determines when to stop iterating — coverage threshold met or marginal returns diminishing. Shared across all campaigns. | | convergence-sensitivity-analysis | Tests conclusion robustness by perturbing parameters and observing rank changes. Shared across scoring, portfolio, and steel-manning campaigns. | | convergence-web-research | Deep web research with full-page extraction — detailed methodology guides, tutorials, implementation references. | | convergence-web-search | Quick web scan to discover relevant pages — methodology references, case studies, best practices for convergence methods. |
Optional, no fixed order; the final leaf is always a sop.
| Campaign | When to use | | --- | --- | | pairwise-ranking | Pairwise Ranking Campaign — produce global rankings through pairwise comparisons and voting aggregation using Bradley-Terry, Elo, TrueSkill, Condorcet, Borda, Schulze methods. |
<!-- END available-tables (generated) -->
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 33,118 | 29,255 | -12% | 1 | 1 | 0% | 6,232 | 7,561 | +21% | 0 | 0 | — |
case-08 | pass→pass | 11,802 | 29,175 | +147% | 1 | 1 | 0% | 2,313 | 7,516 | +225% | 0 | 0 | — |
case-02 | fail→fail | 29,447 | 11,655 | -60% | 1 | 1 | 0% | 6,213 | 1,940 | -69% | 0 | 0 | — |
case-03 | fail→fail | 26,816 | 23,426 | -13% | 1 | 1 | 0% | 5,310 | 6,100 | +15% | 0 | 0 | — |
case-04 | pass→pass | 19,974 | 34,347 | +72% | 1 | 1 | 0% | 3,774 | 7,532 | +100% | 0 | 0 | — |
case-05 | fail→fail | 19,690 | 30,511 | +55% | 1 | 1 | 0% | 4,235 | 7,521 | +78% | 0 | 0 | — |
case-06 | pass→fail | 25,824 | 33,957 | +31% | 1 | 1 | 0% | 2,519 | 7,538 | +199% | 0 | 0 | — |
case-07 | fail→pass | 16,917 | 31,670 | +87% | 1 | 1 | 0% | 2,882 | 7,519 | +161% | 0 | 0 | — |
case-09 | pass→fail | 27,462 | 32,127 | +17% | 1 | 1 | 0% | 5,363 | 7,531 | +40% | 0 | 0 | — |
case-10 | pass→fail | 23,790 | 19,094 | -20% | 1 | 1 | 0% | 5,165 | 3,067 | -41% | 0 | 0 | — |
case-11 | fail→fail | 20,241 | 18,844 | -7% | 1 | 1 | 0% | 3,742 | 4,773 | +28% | 0 | 0 | — |
case-12 | fail→fail | 22,404 | 19,210 | -14% | 1 | 1 | 0% | 5,024 | 5,866 | +17% | 0 | 0 | — |
case-13 | pass→pass | 17,594 | 23,431 | +33% | 1 | 1 | 0% | 3,721 | 7,065 | +90% | 0 | 0 | — |
case-14 | fail→pass | 4,406 | 16,932 | +284% | 1 | 1 | 0% | 767 | 4,962 | +547% | 0 | 0 | — |
case-15 | pass→fail | 24,733 | 31,733 | +28% | 1 | 1 | 0% | 5,357 | 7,527 | +41% | 0 | 0 | — |
case-16 | pass→fail | 22,948 | 26,436 | +15% | 1 | 1 | 0% | 4,544 | 7,523 | +66% | 0 | 0 | — |
case-17 | pass→pass | 17,092 | 29,388 | +72% | 1 | 1 | 0% | 3,288 | 7,522 | +129% | 0 | 0 | — |
case-18 | pass→fail | 25,213 | 28,586 | +13% | 1 | 1 | 0% | 5,197 | 7,513 | +45% | 0 | 0 | — |
case-19 | pass→pass | 25,360 | 33,706 | +33% | 1 | 1 | 0% | 4,706 | 7,508 | +60% | 0 | 0 | — |
case-20 | pass→pass | 15,305 | 29,848 | +95% | 1 | 1 | 0% | 3,126 | 7,536 | +141% | 0 | 0 | — |
case-21 | pass→pass | 7,756 | 22,425 | +189% | 1 | 1 | 0% | 1,595 | 5,034 | +216% | 0 | 0 | — |
case-22 | pass→pass | 12,300 | 15,728 | +28% | 1 | 1 | 0% | 1,841 | 3,811 | +107% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 20 counted toward the lift figure. The other 2 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of -18 percentage points is the difference between those two pass rates over the 20 comparable cases. 7 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.