Catalog
salismt/multi-agent-debate

salismt

multi-agent-debate

Use when tackling a high-stakes, ambiguous, or contested question with no verifiable ground truth and a costly downside — strategy, architecture trade-offs, research synthesis, market or product decisions, investment calls, or any "what should we do about X" where a single confident answer is risky.

v1.0LATEST
NewUpdated Sep 19, 2026

Multi-Agent Debate

Overview

On hard problems the failure mode is a single, fluent, confident answer. This skill replaces that with structured disagreement, evidence grounding, explicit verification, and a resumable record.

Core principle: truth survives the strongest objection — so manufacture that objection on purpose.

When to Use

Use when the question has any of: high stakes, genuine ambiguity, contested expert opinion, no checkable ground truth, or load-bearing assumptions. Symptoms: the honest answer is "it depends"; strategy / architecture / research / market / product / investment calls; decisions you will be held to later.

Calibrate effort to stakes × uncertainty — triage first:

  • Tier 1 — simple or verifiable: answer directly, cite sources, run one sanity check, stop.
  • Tier 2 — hard / contested / high-stakes: run the full protocol below.

Do NOT run the full protocol for lookups, well-specified tasks, or anything with a clear right answer. Debate has real token/latency cost and measurably reduces answer diversity, so invoke it by complexity, not by default.

Set the objective before debating — it controls how hard you push for consensus:

  • Converge → you need one decision. Push toward agreement; the rank-adaptive accelerator in step 4 is allowed.
  • Explore → you need a map of strong, distinct options. Preserve disagreement; do not force consensus.

Debate compresses diversity, so converge only as much as the objective demands.

The Protocol (Tier 2)

  1. Frame & ground. Restate the real question, the decision it informs, success criteria, constraints, and the objective (converge vs. explore). Build an evidence ledger: tag every claim FACT / INFERENCE / ASSUMPTION with its source. Missing data → mark OPEN. Never invent numbers.
  2. Debate as parallel subagents — independent first, then exchange. Dispatch each persona as a separate subagent at the highest reasoning effort available (Claude Code: Task/subagent tool; Codex / OpenCode: parallel agents; reasoning "extra-high" / "high"). Give each a conflicting incentive (e.g. domain operator vs. venture strategist; builder vs. red-team skeptic).
    • Round 1 — isolated: each proposes independently with no visibility of the others (this is what preserves genuine diversity — same-round visibility collapses it).
    • Round 2+ — cross-round exchange: each now sees the others' prior-round positions and attacks the strongest opposing claim with sourced objections. Agreement counts only if it survives the attack.
    • Optional, high-stakes turns: generate 2–N candidate responses and let the judge keep the best (best-of-N) to cut sampling noise.
  3. Judge, don't average. A synthesizer subagent (also maximum effort) scores each contested point against the LLM-as-judge rubric below, records why one side won, and cites the evidence behind each score. Preserve unresolved disputes as OPEN — never force false consensus.
  4. Refine. Revise positions in light of the critique. Bound to 2–3 rounds; stop when a round surfaces no new objection.
    • Optional — converge mode only — rank-adaptive acceleration: each round the judge ranks the positions and prunes or merges the weakest, putting the strongest first. Speeds convergence but trades away diversity; never use it in explore mode.
  5. Verify (chain-of-verification). Generate 5–8 verification questions targeting the load-bearing claims and the top recommendation. Answer each only from sources. Fix every inconsistency. List residual risks.
  6. Decide. Output a ranked recommendation scored on an explicit framework (value × effort × risk, or RICE/ICE), with the winning argument and the strongest surviving objection. Decision-useful: owned, measurable, sequenced.
  7. Persist for resumption. Write a durable record — decisions + rationale, open questions + what evidence closes each, assumptions ledger, changelog — so the next session resumes the analysis instead of restarting it.

Model Assignment

Debate tolerates heterogeneous models; judging does not.

  • Personas — vary the model where you can. Cross-model debate beat both participants debating alone (Du et al.), and weaker models gain proportionally more from the protocol than frontier ones.
  • Judge + verification — use the strongest model available and keep it fixed. Judges favor their own model family, so drawing the judge from one debater's family biases the verdict (Liang et al.).
  • On a weak base model, cut the round count and tighten the rubric. Debate amplifies the reasoning a model already has; it does not substitute for it.

Judge Rubric (evidence-based LLM-as-judge)

Score every contested claim and the final recommendation 1–5 on each criterion. A score is valid only if the judge cites the specific evidence — an evidence-ledger row, a source, or a numeric check — that justifies it; bare numbers and vague justifications are themselves a fail. Anything scoring <4 goes back for another refine round (step 4) or is marked OPEN.

Criterion Scores 5 when…
Grounding every factual claim traces to a source in the evidence ledger; no unsourced numbers
Completeness covers the decision's material options and both personas' strongest points
Decision-usefulness recommendation is ranked, owned, measurable, and sequenced
Specificity concrete to the actual context, not generic boilerplate
Internal consistency no contradiction with other accepted claims or stated constraints
Robustness survives its strongest stated objection and the verification questions (step 5)

Quick Reference

Phase Output Guardrail
Frame & ground Question + objective + evidence ledger No claim without a source tag; pick converge vs. explore
Debate Competing positions Parallel subagents, max effort, isolated round 1 then exchange
Judge Scored, adjudicated synthesis Score on the evidence-based rubric; cite evidence per score
Refine Revised positions Stop when no new objection; rank-adaptive prune only in converge mode
Verify Checked claims + risks Answer only from sources
Decide Ranked, scored recommendation Owned + measurable
Persist Resumable record Decisions, open Qs, assumptions, changelog

Common Mistakes

Rationalization Reality
"The two views agreed fast" They aren't divergent. Give them conflicting incentives and a red-team mandate.
"One agent can play both sides" Same context bleeds the positions together. Run each persona as a separate subagent at max effort.
"Let them see each other from round 1" Same-round visibility collapses independence. Keep round 1 isolated; exchange from round 2.
"I'll just pick the better-written side" Score both on the evidence-based rubric and cite evidence per score, or the verdict is vibes.
"Converge to a clean answer" (goal was explore) You destroy the option diversity you needed. Match consensus pressure to the objective.
"No source, so I'll estimate" Mark it OPEN. An invented number poisons every downstream conclusion.
"Verification is overkill" The load-bearing claim is the one most likely wrong — verify it first.
"I'll skip the record" Then it can't be resumed and the next agent repeats the work.

Red Flags — STOP

  • Round 1 wasn't isolated (peers visible too early) → diversity already lost.
  • Personas converging the moment they exchange → not adversarial enough.
  • Forcing consensus when the objective was explore.
  • Any unsourced figure in the output → ground it or mark OPEN.
  • A recommendation with no owner, metric, or surviving objection → not decision-useful.
  • About to conclude with no verification pass on the top claim.

Why This Works

Composes multi-agent debate (counters Degeneration-of-Thought), divergent-persona role-play, Self-Refine, Chain-of-Verification, and evidence-grounded LLM-as-judge evaluation. The protocol knobs — round-1 isolation, cross-round exchange topology, optional rank-adaptive scheduling, and complexity-triggered invocation — follow controlled MAD-protocol findings that interaction and convergence trade off, and that debate reduces diversity. See references.md.

Files1
1 files · 6.5 KB

Select a file to preview

Overall Score

89/100

Grade

A

Excellent

Grades are signals, not a certification. Always review a skill yourself before use.

Safety

92

Quality

88

Clarity

87

Completeness

85

Summary

This skill provides a structured protocol for multi-agent debate on hard, ambiguous, or high-stakes decisions with no verifiable ground truth. It synthesizes controlled research on multi-agent debate, chain-of-verification, self-refinement, and evidence-grounded LLM judging to replace single confident answers with systematic disagreement, evidence grounding, and resumable records. The protocol scales effort to complexity and supports both convergent and exploratory objectives.

Detected Capabilities

multi-turn reasoningsubagent dispatchstructured evidence trackingchain-of-verificationevidence-based rubric scoringdecision synthesis and ranking

Trigger Keywords

Phrases that agents use to match this skill to user intent.

resolve contested decisiondebate high-stakes choicestrategy trade-off analysissynthesize expert disagreementaudit top recommendationresearch synthesis with conflict

Referenced Domains

External domains referenced in skill content, detected by static analysis.

aclanthology.orgarxiv.org

Use Cases

  • Resolving contested strategy or architecture decisions with competing viewpoints
  • Synthesizing research or market analysis when expert opinion is divided
  • Evaluating investment or product decisions where ground truth is unknowable before execution
  • Generating explored option maps for decisions where multiple strong paths exist
  • Auditing high-stakes recommendations to surface hidden assumptions and load-bearing claims
  • Building resumable decision records that preserve reasoning and open questions across sessions

Quality Notes

  • Strong evidence grounding: every step references published research with citation and specific results; references.md provides transparent basis for protocol decisions
  • Practical guardrails: explicit red flags, common-mistakes table, and triage rules prevent misuse and over-application of expensive debate
  • Clear objective framing: converge-vs-explore knob and tier-based triage ensure alignment between effort and decision stakes
  • Robust verification: chain-of-verification step and evidence-ledger discipline counter hallucination and unsourced claims
  • Handles honest limits: acknowledges Smit et al. 2023 finding that MAD does not outperform simpler strategies by default; presents debate as a targeted tool with hyperparameter sensitivity
  • Heterogeneous model support: explicitly covers cross-model debate and cost-benefit tradeoffs (personas on cheaper models, judge on strongest)
  • Resumable records: persistence requirement (step 7) enables incremental analysis and prevents work duplication
  • Transparent research basis: references are not cherry-picked; includes counterweight (Smit et al.) alongside supportive results; caveat on Zargari Marandi 2026 (domain-specific, proxy metrics)
  • Well-structured sections: hierarchy from overview → when-to-use → full protocol → rubric → reference is logical and easy to navigate
  • Actionable error diagnosis: common-mistakes and red-flags tables give agents concrete patterns to detect and correct
Model: claude-haiku-4-5-20251001Analyzed: Sep 19, 2026

Reviews

Add this skill to your library to leave a review.

No reviews yet

Be the first to share your experience.

Use salismt/multi-agent-debate in your dev environment

Command Palette

Search for a command to run...