Which integrations are available for this server?

Allows running automated debate sessions using OpenAI's API as the backend for generating positions and challenges.

1. Click on "Deploy Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Ploidy debate: should we adopt a monorepo?" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

Ploidy

by heznpc

Overview Schema Source Related Servers Discussions Admin

Python

Hybrid

Ploidy

Context-asymmetric structured debate for difficult decisions.

Ploidy gives one side of a debate project context and keeps the other side fresh, then records positions before either side can anchor on the other. It ships as a Python 3.11+ MCP server with SQLite persistence.

Install

pip install ploidy

Add the API extra only when the server itself will generate an automatic debate:

pip install "ploidy[api]"

Related MCP server: consensus-mcp

Recommended local setup: stdio

stdio is the default transport. One MCP client starts one server on demand, so there is no port or background process to manage.

{
  "mcpServers": {
    "ploidy": {
      "type": "stdio",
      "command": "python3",
      "args": ["-m", "ploidy"]
    }
  }
}

The primary tool is debate:

mode="solo": the caller supplies independently generated Deep and Fresh positions. No external model API is required.
mode="auto": the server generates both sides through an OpenAI-compatible API. Non-empty context_documents are required so the Deep and Fresh sides are actually asymmetric.

Example auto call:

await debate(
    prompt="Should we split the ingestion service?",
    mode="auto",
    context_documents=["Current topology, incident history, and constraints…"],
    context_sources=["architecture-notes"],
)

Configure the API backend before using auto mode:

export PLOIDY_API_BASE_URL=https://api.openai.com/v1
export PLOIDY_API_KEY=...
export PLOIDY_API_MODEL=your-model

Shared server: Streamable HTTP

Use HTTP when multiple MCP clients must share one debate or when Ploidy is deployed as a service:

PLOIDY_TRANSPORT=streamable-http python3 -m ploidy
export PLOIDY_URL=http://127.0.0.1:8765

Configure each client with ${PLOIDY_URL}/mcp (substitute the value; JSON does not expand shell variables):

{
  "mcpServers": {
    "ploidy": {
      "type": "streamable-http",
      "url": "http://127.0.0.1:8765/mcp"
    }
  }
}

PLOIDY_URL is the base URL used by ploidy-ask and other HTTP clients. The MCP endpoint is /mcp, and the live HTTP/SSE endpoint is /v1/debate/stream.

MCP tools

Ploidy 0.4.0 exposes exactly 13 tools. New integrations should use the unified debate tool. The other 12 remain available for explicit two-client, phase-by-phase, history, cancellation, and HITL workflows.

Tool	Purpose
`debate`	Unified `auto` or `solo` debate
`debate_start`	Create a Deep session
`debate_join`	Join as Fresh or Semi-Fresh
`debate_position`	Submit an independent position
`debate_challenge`	Submit a semantic challenge
`debate_converge`	Produce the convergence result
`debate_cancel`	Cancel an active debate
`debate_delete`	Permanently delete a debate
`debate_status`	Read current state
`debate_history`	List visible debate history
`debate_solo`	Legacy alias for caller-supplied debate
`debate_auto`	Legacy alias for API-generated debate
`debate_review`	Approve, override, or reject a paused auto debate

Service configuration

Common environment variables:

PLOIDY_TRANSPORT=stdio             # stdio (default) or streamable-http
PLOIDY_PORT=8765
PLOIDY_DB_PATH=~/.ploidy/ploidy.db
PLOIDY_MAX_CONTEXT_TOKENS=20000    # recommended for hosted auto mode
PLOIDY_RATE_CAPACITY=20
PLOIDY_RATE_PER_SEC=1
PLOIDY_RETENTION_DAYS=30

Authentication modes and their limits:

Admission-controlled self-hosting: PLOIDY_AUTH_MODE=bearer with PLOIDY_TOKENS='{"token":"tenant"}'.
OAuth interoperability: PLOIDY_AUTH_MODE=oauth and an exact public PLOIDY_OAUTH_ISSUER; v0.4.0 auto-approves clients and therefore is not user login or public admission control.
Migration only: PLOIDY_AUTH_MODE=both adds the static-token fallback.

Bearer mode without a configured token is intentionally unauthenticated and is suitable only for loopback development. Public services need a private ingress/gateway or controlled static tokens until OAuth gains resource-owner login and consent.

Container

The published image is immutable by release version:

docker run --rm \
  -p 127.0.0.1:8765:8765 \
  -v ploidy-data:/data \
  ghcr.io/heznpc/ploidy:0.4.0

Or run the checked-in Compose configuration:

docker compose up --build

The image defaults to streamable-http and contains the API, dashboard, metrics, Redis, and CLI runtime extras.

Documentation

License

MIT

Available Tools

13 tools

debateA

Destructive

Run a context-asymmetric debate in a single call.

One canonical entry point. Pick a mode:

auto (default): Ploidy generates both sides via the configured OpenAI-compatible API endpoint and returns the convergence result. Requires non-empty context_documents plus a configured API endpoint. HITL pause/resume is available via pause_at; resume with the legacy debate_review tool.
solo: you supply both positions (and optionally both challenges) and Ploidy persists + converges them. No external API key needed — the recommended single-terminal flow when the caller (e.g. Claude Code itself) writes both sides locally.

The older 12-tool surface (debate_start / debate_join / debate_position / ...) remains available for two-terminal and HITL workflows but is now deprecated; new integrations should use this tool.

Args: prompt: The decision question to debate. mode: "auto" or "solo". deep_position: (solo only) Deep side's stance. fresh_position: (solo only) Fresh side's stance. deep_challenge: (solo only) Optional deep-side critique. fresh_challenge: (solo only) Optional fresh-side critique. deep_label / fresh_label: (solo) display labels for roles. fresh_role: (auto) "fresh" or "semi_fresh". delivery_mode: (auto + semi_fresh) "passive", "active", or "selective". pause_at: (auto) Optional "challenge" or "convergence" HITL pause point. deep_n / fresh_n: (auto) Ploidy level per side. effort: (auto) "low", "medium", "high", "max". injection_mode: (auto) Context formatting mode. context_pct: (auto) Percentage of context to retain. language: (auto) Output language code. deep_model / fresh_model: (auto) Model override. Supplying either applies it to both sides; if both are supplied they must match. context_documents: Documents attached only to the Deep side. Required and non-empty in auto mode; optional provenance in solo mode. context_sources: Optional provenance labels, one per context document. blocked_sources: Optional strings that must not appear in source labels or context documents. target_lease: Optional target identifier that pins this debate's scope. allowed_sources: Optional source-label allowlist; defaults to target_lease when a target lease is set.

Returns: Convergence result dict, or paused state (auto + pause_at).

Inputs: prompt (required), mode, deep_n, effort, +20 more.

debate_autoA

Destructive