llm-sparring

Overview Schema Source Related Servers Discussions Admin

Synthesize with GPT

synthesize

Read-only

Generates an independent GPT answer and a blind comparison with your answer to surface agreements, conflicts, unique points, and a merged response.

Instructions

Get GPT's independent answer to a question, then a blind comparison that merges it with your answer: agreements, conflicts (and which side is better supported), points only one answer has, and a merged answer. GPT answers first without seeing your answer; a second, blind call compares the two as Answer A/B in random order. Call this for questions where an independent second answer adds value: factual or technical questions with real stakes, estimates, recommendations, or when the user asks what another model thinks. Write your complete answer first and pass it as claude_answer. This makes two GPT calls, so it is slower than the other tools.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "question": "<question>",
      "claude_answer": "<claude_answer>"
    },
    "name": "synthesize"
  }
}

Input Schema

Name	Required	Description	Default
`question`	Yes	The question, with any context GPT needs (it sees nothing else).
`claude_answer`	Yes	Your (Claude's) complete answer.

Output Schema

Name	Required	Description	Default
`model`	Yes
`side_a`	Yes
`gpt_answer`	Yes
`merge_failed`	Yes
`merge_parsed`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.3.0Nov 10, 2026

TDQS

A4.8/5.0

Behavior5/5

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

Annotations only declare read-only/open-world/non-idempotent; the description adds genuinely new behavioral facts: GPT answers first without seeing your answer, the comparison is a second blind call with A/B in random order, and the tool costs two GPT calls (latency). That is exactly the kind of detail annotations cannot convey.

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

Conciseness4/5

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

Front-loads the purpose, then the conditions, then the cost caveat; each part is useful. It is slightly long and re-explains the blind-comparison mechanics that the opening sentence already implies, but there is no filler.

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?

An output schema exists, yet the description still summarizes the returned components (agreements, conflicts with support assessment, unique points, merged answer), covering both the call contract and the result shape. Nothing an agent needs to invoke it correctly is missing.

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

Parameters4/5

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

Schema coverage is 100% and both parameters are already documented, so the baseline is 3. The description adds ordering semantics — the answer must be written first and passed as claude_answer — which is workflow meaning beyond the schema's field descriptions.

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?

Names a specific verb and resource — get GPT's independent answer plus a blind comparison that produces agreements, conflicts, one-sided points, and a merged answer. The output shape and the two-call mechanism are stated concretely, so an agent knows exactly what it will receive.

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?

Explicitly scopes usage: factual/technical questions with real stakes, estimates, recommendations, or when the user asks what another model thinks. It also states the prerequisite workflow — write your complete answer first and pass it as claude_answer — and warns this is slower than the other tools, which lets an agent trade off latency against value.

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

Deploy Server