Skip to main content
Glama

End-of-task model review

review_model_choice

Reviews the model used after a substantial task and recommends a lighter or stronger model for next time when the result suggests a mismatch.

Instructions

Call at the END of a substantial task. Says whether a lighter (or stronger) model would have been better next time. Show the returned note to the user only if change is 'up' or 'down'.

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": {
      "task": "<task>",
      "model_used": "<model_used>"
    },
    "name": "review_model_choice"
  }
}

Input Schema

NameRequiredDescriptionDefault
taskYesWhat was done
model_usedYesModel that did the task, e.g. 'Opus 5.5'
output_sizeNo
result_was_goodNoFalse if the result fell short

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv1.1.0

TDQS

A3.8/5.0
Behavior3/5

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

No annotations exist, so the description has to cover the behavioral burden itself. It does disclose the call timing, that a 'note' is returned, and that the note carries a 'change' value that can be 'up' or 'down'; but it never says whether the call mutates state, whether it costs a model invocation, or how quickly it returns. The disclosure is partial and functional rather than complete.

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?

Three short sentences, with the most decision-relevant information (when to call) front-loaded. Nothing is redundant and nothing needs trimming.

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?

The tool is simple, with 4 flat parameters and no output schema, and the definition covers the trigger, the outcome, and a display rule. The main gap is the return shape: an agent knows a 'note' with a 'change' field exists but not its full structure or the other possible values of 'change'.

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 75%: task, model_used, and result_was_good are documented in the schema, while output_size has only an enum. The description adds no meaning to any parameter, leaving the ambiguity of output_size and the interaction between result_was_good and the returned verdict unaddressed. That lands at the baseline for near-high coverage.

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?

The definition states a specific action and its trigger: 'Call at the END of a substantial task' and 'Says whether a lighter (or stronger) model would have been better next time.' That is a concrete verb+resource an agent can match to intent. It never names or contrasts with the siblings recommend_model or list_models, so the differentiation is left implicit.

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?

'Call at the END of a substantial task' gives an explicit timing condition, and the third sentence adds a downstream rule for what to do with the result ('Show the returned note to the user only if change is up or down'). What's missing is a when-not clause, e.g. for trivial tasks, or a pointer to recommend_model as the alternative for pre-task selection.

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

Deploy Server

Other Tools