# Iterative Convergence Round

> Execute one full Delphi round — collect judgments, distribute anonymous feedback, measure consensus, decide whether to continue.

- Skill: `yogsoth-ai/iterative-convergence-round` (Agent Skill)
- Install (CLI): `npx skillmds@latest add yogsoth-ai/iterative-convergence-round`
- Raw SKILL.md: https://api.skillmd.com/api/skills/yogsoth-ai/iterative-convergence-round/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: yogsoth-ai (https://skillmd.com/u/yogsoth-ai)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/yogsoth-ai/iterative-convergence-round

---


# Iterative Convergence Round

Execute one complete iteration of the Delphi convergence cycle: collect independent judgments from all perspectives, distribute anonymized feedback showing the group distribution, allow revision, measure consensus level, and decide whether another round is needed.

## Stages

1. **Collect** — Run `judgment-collection` to gather independent ratings/estimates from each perspective
2. **Feedback** — Run `feedback-distribution` to create anonymized summary of group responses
3. **Measure** — Run `consensus-measurement` to compute agreement level (IQR, % agreement, Kendall's W)
4. **Decide** — Run `round-decision` to determine continue/stop based on threshold and stability

## Available SOPs

| SOP | Role in Tactic |
|-----|---------------|
| judgment-collection | Gather independent judgments from all perspectives |
| feedback-distribution | Create and distribute anonymized feedback report |
| consensus-measurement | Compute consensus score using appropriate method |
| round-decision | Determine whether to run another round or stop |

## Execution Guidance

- Judgments MUST be collected independently (no cross-contamination)
- Feedback MUST be anonymized (no attribution to specific perspectives)
- Consensus measurement method should match data type (IQR for continuous, % for categorical)
- Stop conditions: consensus threshold met OR stability detected OR max rounds reached
- If stopping without consensus, document which items remain unresolved

## Minimum Yield

- Consensus items list (items that reached consensus)
- Non-consensus items (items that did not converge, with reasons)
- Round log (round-by-round progression of scores)

<!-- BEGIN available-tables (generated) -->

## Available SOPs

Optional, no fixed order; the final leaf is always a sop.

| SOP | When to use |
| --- | --- |
| consensus-measurement | Compute consensus score from collected judgments using the appropriate statistical method. |
| feedback-distribution | Create anonymized feedback report summarizing group judgment distribution for a given round. |
| judgment-collection | Collect independent judgments from all perspectives on a given question. |
| round-decision | Decide whether to continue iterating or stop based on consensus score, round number, and stability. |

<!-- END available-tables (generated) -->