Skip to main content
Glama

debate

Destructive

Run a single-call, context-asymmetric deliberation that pits opposing positions against each other to minimize confirmation bias and converge on a balanced outcome.

Instructions

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).

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
modeNoauto
deep_nNo
effortNohigh
promptYes
fresh_nNo
languageNoen
pause_atNo
deep_labelNoDeep
deep_modelNo
fresh_roleNofresh
context_pctNo
fresh_labelNoFresh
fresh_modelNo
target_leaseNo
deep_positionNo
delivery_modeNonone
deep_challengeNo
fresh_positionNo
injection_modeNoraw
allowed_sourcesNo
blocked_sourcesNo
context_sourcesNo
fresh_challengeNo
context_documentsNo
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only provide readOnlyHint=false, idempotentHint=false, destructiveHint=true. The description adds substantial behavioral context beyond these annotations: it explains that auto mode generates both sides via the API, solo mode persists and converges user-supplied positions without external API keys, and pause_at creates HITL pause points that require resuming with the legacy debate_review tool. This meaningfully characterizes the tool's runtime behavior and requirements.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is long but well justified for a 24-parameter tool. It front-loads the core purpose in the first sentence, uses clear bullet-style sections for modes, deprecation, and Args, and every sentence contributes useful operational detail. The structure mirrors the schema's parameter list while adding essential mode-specific context, making it easy to scan.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (24 parameters, no output schema, and minimal annotations), the description is remarkably complete. It covers both operating modes, prerequisites, HITL pause/resume, deprecated siblings, parameter applicability, and the return type (convergence result dict or paused state). The only missing piece would be a detailed output schema, but the description's Returns line offers a sufficient hint.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

With a schema description coverage of 0%, the description fully compensates by listing all 24 parameters with mode annotations (e.g., '(solo only)', '(auto)') and cross-parameter constraints. It clarifies that deep_position/fresh_position are solo-only, deep_model/fresh_model apply to both sides when either is supplied and must match if both are given, and context_documents must be non-empty in auto mode. This goes far beyond the bare schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a specific verb+resource statement: 'Run a context-asymmetric debate in a single call' and immediately identifies this as the 'One canonical entry point.' It explicitly distinguishes itself from the 12-tool legacy surface by declaring that surface deprecated and recommending this tool for new integrations, which resolves ambiguity among siblings.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides explicit when-to-use guidance: it names the deprecated debate_start/join/position tools as alternatives for two-terminal/HITL workflows, explains the auto mode requires a configured API endpoint and non-empty context_documents, and highlights solo mode as the recommended single-terminal flow. It also notes how to resume HITL pauses with debate_review, giving clear decision criteria.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/heznpc/PLOIDY'

If you have feedback or need assistance with the MCP directory API, please join our Discord server