openai-decisions-mcp

Overview Schema Source Related Servers Discussions Admin

evaluate_batch

Read-only

Evaluate up to 20 named text, file, image, or PDF items against shared predicate, choice, or score questions and return ordered per-item results or errors.

Instructions

Apply common typed questions independently to up to 20 named items, each containing text or file/image/PDF sources. At most 3 evaluations run concurrently. Returns per-item results/errors in input order; never replays successful items. Paid requests: one per valid item. Caller chooses the criteria and subsequent selection.

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": {
      "items": "<items>",
      "questions": "<questions>"
    },
    "name": "evaluate_batch"
  }
}

Input Schema

Name	Required	Description
`full`	No
`items`	Yes	Prefer a native JSON array. A JSON-encoded array string is also accepted for client compatibility; identical validation applies.
`questions`	Yes	Prefer a native JSON array. A JSON-encoded array string is also accepted for client compatibility; identical validation applies.

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.2.1Oct 29, 2026

TDQS

A3.7/5.0

Behavior4/5

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

Beyond the annotations, it discloses a concurrency ceiling of 3, per-item results/errors returned in input order, the no-replay guarantee for successful items on retry, and that cost is one paid request per valid item. That is real operational context an agent could not get from readOnlyHint/idempotentHint alone. What remains unexplained is what happens to partially failed batches and what 'full' changes.

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 core operation and packs concurrency, return order, retry and billing into two compact sentences with little waste. Some phrasing is telegraphic, notably the closing sentence about the caller choosing criteria and selection, which reads as an inside reference rather than actionable guidance.

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

Completeness4/5

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

With no output schema, the description carries the burden of describing returns and does so: per-item results and errors in input order. Combined with the billing and concurrency disclosure, an agent has enough to call it correctly, though the meaning of 'full' and per-item failure semantics are gaps.

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

Parameters3/5

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

Schema coverage is 67%, and the description adds the source types (text, file, image, PDF) and the 'typed questions' concept that maps to the predicate/choice/score union. However, the third parameter 'full' is never mentioned and no guidance is given on how many questions to attach or how ids are used in results.

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

Purpose4/5

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

States a specific verb and resource: apply common typed questions independently to up to 20 named items whose sources are text or file/image/PDF. Scope and batch limits are explicit, so an agent can picture the operation immediately. It does not differentiate itself from the siblings decide or find_code, so it stops short of a 5.

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

Usage Guidelines3/5

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

Usage is implied by the batch shape and item cap, but there is no explicit statement of when to pick this over decide (single-item) or when not to use it. 'Caller chooses the criteria and subsequent selection' gestures at the workflow without defining it.

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

Deploy Server