Skip to main content
Glama

Evaluate response quality

evaluate_response_quality
Read-onlyIdempotent

Scores a prompt and model reply with a local heuristic rubric, returning JSON quality_score, passed, and criteria without calling a model.

Instructions

Score a prompt and reply with a local heuristic rubric. Does not call a model. Use run_calibration_test to generate replies first. Read-only. Returns JSON with quality_score, passed and criteria.

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": {
      "prompt": "<prompt>",
      "response": "<response>"
    },
    "name": "evaluate_response_quality"
  }
}

Input Schema

NameRequiredDescriptionDefault
promptYesOriginal prompt text.
categoryNoRubric category such as general, analysis or coding. Default general.general
responseYesModel reply to score.

Output Schema

NameRequiredDescriptionDefault
resultYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changedv6.2.1
    • addedInput schema / properties / category / description
      Added value: +"Rubric category such as general, analysis or coding. Default general."
    • addedInput schema / properties / prompt / description
      Added value: +"Original prompt text."
    • addedInput schema / properties / response / description
      Added value: +"Model reply to score."
  2. First observedv5.0.0

TDQS

A4/5.0
Behavior4/5

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

Annotations already cover readOnlyHint, idempotentHint, destructiveHint and openWorldHint, yet the description still adds real information: the scoring is local/deterministic and makes no model call (no API cost, latency, or provider dependency), and it names the returned fields quality_score, passed and criteria.

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?

Four terse sentences, zero filler: what it does, what it does not do, the prerequisite step, and the return payload. The key disambiguator ('Does not call a model') is placed immediately after the purpose statement.

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?

Annotations and an output schema carry most of the burden, and the description supplies the workflow prerequisite plus the shape of the result. It never explains what the rubric actually measures or how quality_score and passed relate to criteria, which is the one remaining gap for a scoring tool.

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 description coverage is 100%, so the schema fully documents prompt, response and category (including the 'general, analysis or coding' rubric values). The description adds no further meaning about how category changes scoring, so the baseline 3 applies.

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?

Specific verb+resource: 'Score a prompt and reply with a local heuristic rubric.' The qualifier 'local' plus 'Does not call a model' distinguishes it from model-invoking siblings such as chat_with_model or compare_model_responses, and it explicitly positions run_calibration_test as the upstream step.

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

Usage Guidelines4/5

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

'Use run_calibration_test to generate replies first' gives an explicit prerequisite and routing to the sibling that produces the input. It stops short of stating when NOT to use it (e.g., scoring already-stored conversations vs. live replies), so no exclusions are covered.

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