Skip to main content
Glama

Create a poll

create_poll

Create a poll from a definition: questions and their wordings, the models to ask (provider and model from list_models), samples per cell, search depth, schedule and dates. It is saved as a draft and costs nothing; call estimate_poll for its cost, then start_poll or run_poll_now. A test key makes a test poll, which runs on a stub model and charges nothing. Send an Idempotency-Key so a retry cannot make two polls. (POST /api/v1/polls)

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": {
      "title": "<title>",
      "schedule": "<schedule>",
      "samples_per_cell": "<samples_per_cell>",
      "questions": "<questions>",
      "panel": "<panel>"
    },
    "name": "create_poll"
  }
}

Input Schema

NameRequiredDescriptionDefault
panelYes1 to 30 models to ask, each once.
titleYesThe poll's title.
deliveryNorealtime (the default), or batch: scheduled runs ask the models that can be batched (Anthropic) through the providers' batch APIs, cheaper, with answers arriving up to 2 hours after the run starts; run_poll_now is always real time. OpenAI and Google models are asked at their flex tier, the same price, in real time, either way. Left out: unchanged.
end_dateNoThe last day it runs, at most 90 days after start_date.
run_timeNoThe time of day it runs, HH:MM in time_zone; null runs from the moment it is started.
scheduleYesonce, daily or weekly.
questionsYes1 to 50 questions, in order.
time_zoneNoThe IANA time zone of its dates and run time.America/Anchorage
index_nameNoA name for the composite of the poll's Better-Worse index questions, such as "AI Economy Index"; empty for "Composite Better-Worse index". Shown only when two or more questions have the index.
start_dateNoThe first day it runs, YYYY-MM-DD in time_zone; needs run_time.
change_noteNoWhy this version differs from the last, kept in the poll's history.
descriptionNoWhat the poll is for, or empty.
run_weekdayNoWeekly polls: the day it runs, 0 = Sunday.
search_depthNoSearches allowed per answer, for every model with search on.
restored_fromNoThe earlier version this one was started from, or null.
system_promptNoThe system prompt every wording is asked under unless it has its own; empty for none.
sample_percentNoThe share of the poll's wordings each scheduled run asks, 1 to 100 percent; every wording is asked once before any is asked again. Left out: unchanged (100 for a new poll). run_poll_now asks this share with share=true, every wording otherwise.
samples_per_cellYesHow many times each model is asked each wording per run, 1 to 50: more samples, narrower ranges.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.8/5.0
Behavior5/5

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

Annotations declare a non-read-only, non-destructive, non-idempotent mutation, and the description adds the context that matters: the result is only a draft, no charge is incurred, a test key produces a stub-model/no-cost poll, and an Idempotency-Key is needed because the operation itself is not idempotent. This is consistent with idempotentHint=false rather than contradicting it.

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 dense sentences, each carrying distinct information: what is created, its draft/no-cost state, the test-key variant, and the idempotency requirement. The purpose is front-loaded and nothing is filler.

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?

For an 18-parameter creation tool with no output schema, the description covers cost, lifecycle, test mode and retry safety well. Its main omission is what the call returns (e.g., the new poll's id) and whether required fields like panel/schedule are validated up front, which an agent would want before chaining into estimate_poll.

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%, so the baseline is 3, but the description adds navigational meaning by pointing at list_models as the source of valid provider/model values and by grouping parameters conceptually (wordings, samples per cell, search depth, schedule, dates). It does not explain every field, but the schema already does.

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?

States a specific verb and resource ('Create a poll from a definition') and enumerates what the definition contains: questions and wordings, models, samples per cell, search depth, schedule and dates. It is clearly distinguishable from the sibling tools it names (estimate_poll, start_poll, run_poll_now, list_models).

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?

Gives an explicit lifecycle: create saves a draft that costs nothing, then call estimate_poll for cost, then start_poll or run_poll_now. It also states the conditions that change behavior (test key → test poll on a stub model) and the retry safeguard, so an agent knows exactly when and how to use it versus its siblings.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.