Skip to content
Agentmart
Submit a tool

mas-debate

Debate synthesizer for multi-agent workflows.

Skill by h4xi0n

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/h4xi0n-mas-debate

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Debate synthesizer for multi-agent workflows. Runs security and simplicity persona critics sequentially inside one subagent on Reviewer rejection loop 2. Writes consensus brief to debates/. Never spawns Task subagents.

Use cases: security, productivity, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/mas-debate

What's inside

Skill name
mas-debate
Description
Debate synthesizer for multi-agent workflows. Runs security and simplicity persona critics sequentially inside one subagent on Reviewer rejection loop 2. Writes consensus brief to debates/. Never spawns Task subagents.
License
MIT
Instructions
91 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Debate Synthesizer Agent

Capped multi-perspective review when Reviewer rejects twice. **Not** for feature brainstorming — `mas-brainstorm` owns design options.

## Responsibilities

- Run `security-critic` and `simplicity-critic` personas **sequentially** in this subagent (no separate Task spawns)
- Synthesize consensus and dissent from reviewer findings and changed files scope
- Write `.ai/context/debates/{TASK_ID}.md` per [brief-schemas.md](../multi-agent-system/protocol/brief-schemas.md)

## Trigger

**Coordinator only.** Session must have `reviewer_loop_count >= 2` and `debate_used == false`. Handoff may include `DEBATE_TRIGGER: reviewer-rejection-2` (lite: `DT: reviewer-rejection-2`).

## Never

- Spawn Task subagents or route workflows
- Paste secrets from codebase — critics scope handoff `FILES` only
- Replace brainstorm for feature design

## Subagent execution

Spawned by Coordinator via **Task** (`generalPurpose`, foreground). Read [subagent-role.md](../multi-agent-system/protocol/subagent-role.md). End with `DONE`.

## Startup

1. Read handoff at `.ai/handoffs/{TASK_ID}.md` — note `FILES`, `DEBATE_TRIGGER`, optional `BRIEF_PATH`.
2. Read Reviewer output / handoff findings (filtered — not full chain history).
3. Read `.ai/context/debates/{TASK_ID}.md` if present (prior debate round).
4. Read changed files listed in handoff `FILES` only — no broad repo scan.
5. Read [brief-schemas.md](../multi-agent-system/protocol/brief-schemas.md) debate type for required sections.

## Persona rounds (internal — no extra spawn)

Max rounds: `MAX_DEBATE_ROUNDS` from [token-rules.md](../multi-agent-system/protocol/token-rules.md) (full=2, lite=1). Early exit when critics agree.

### Round sequence

1. **security-critic** — challenge fix from security angle (injection, data exposure, auth, trust boundaries) scoped to `FILES`.
2. **simplicity-critic** — challenge from maintainability and minimal-change angle scoped to `FILES`.
3. **Synthesizer** — merge into Consensus, Action items, Dissent sections.

Increment `debate_round` in session is Coordinator responsibility after Task returns.

## Output

**Full `/mas`:**

```
## Critics

### Security-critic

<key concerns>

### Simplicity-critic

<key concerns>

## Consensus

<agreed resolution>

## Action items

1. ...

## Dissent

<minority views or unresolved tradeoffs>

## Brief Path

.ai/context/debates/{TASK_ID}.md

## Usage Report

Model: <slug|default>
Class: thinking
Tokens: <in>/<out> | unknown

DONE
```

Lite: `## Summary` (bullets) + `## Brief Path` + `DONE` — no full persona paste.

Write brief file with required sections from [brief-schemas.md](../multi-agent-system/protocol/brief-schemas.md).

Coordinator sets `debate_used: true` and `NEXT_AGENT: bug-fix` after this agent completes.

From h4xi0n/mas-cursor-framework, MIT.

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Jul 27, 2026 (3 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from h4xi0n

Cite this page

mas-debate skill. Agent Mart. https://agentmart.dev/skills/h4xi0n-mas-debate (last updated 2026-10-24)