Skip to main content
Glama
murtazox04

Cue

by murtazox04
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Remote

request_approval

Ask reviewers to approve or reject an action before you take it, such as deleting data or spending money. Poll the request status and act only when approved.

Instructions

Ask people to approve or reject an action you are about to take outside Cue (deleting data, spending money, contacting many people), before you take it. Not for informing anyone: use send_event or send_message. Needs the messages:write scope.

Cue reports a cue.approval_requested event per reviewer, which the server's rules turn into a notification with a one-time review link; only people decide, no tool can. Returns the request: id, status (pending), title, body, reviewers, importance, metadata, expires_at and created_at. Then poll get_approval with the id and act only on approved. Agent keys may name only their agent's configured reviewers (403 otherwise) and have a limit of open requests (429, 10 by default).

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
bodyNoWhat the reviewer needs to decide (what, why, impact), up to 4,000 characters. Defaults to empty.
titleYesWhat needs a decision, ideally a yes/no question such as `Refund order 9?`; 1-200 characters.
metadataNoJSON object returned unchanged with the request and its decision, e.g. `{"order": "9"}`, to tie it back to your work; at most 16 KB. Defaults to {}.
reviewersYesExternal ids of the 1-10 people who may decide; the first decision settles the request.
importanceNoHow urgently reviewers are notified; defaults to `high`. An agent's ceiling lowers anything above it.high
expires_in_minutesNoMinutes until an undecided request expires, which means no: 1 to 10080 (7 days), smaller values become 1. Defaults to 60.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed6 schema fields changedv0.1.1
    • addedInput schema / properties / body / description
      Added value: +"What the reviewer needs to decide (what, why, impact), up to 4,000 characters. Defaults to empty."
    • addedInput schema / properties / expires_in_minutes / description
      Added value: +"Minutes until an undecided request expires, which means no: 1 to 10080 (7 days), smaller values become 1. Defaults to 60."
    • addedInput schema / properties / importance / description
      Added value: +"How urgently reviewers are notified; defaults to `high`. An agent's ceiling lowers anything above it."
    • addedInput schema / properties / metadata / description
      Added value: +"JSON object returned unchanged with the request and its decision, e.g. `{\"order\": \"9\"}`, to tie it back to your work; at most 16 KB. Defaults to {}."
    • addedInput schema / properties / reviewers / description
      Added value: +"External ids of the 1-10 people who may decide; the first decision settles the request."
    • addedInput schema / properties / title / description
      Added value: +"What needs a decision, ideally a yes/no question such as `Refund order 9?`; 1-200 characters."
  2. First observedv0.1.0

TDQS

A4.8/5.0
Behavior5/5

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

Annotations are minimal (write, open-world, non-destructive), and the description carries the real burden: messages:write scope, a cue.approval_requested event per reviewer, a one-time review link, humans-only decision, 403 for unconfigured reviewers, and a 429 limit of ~10 open requests. This is exactly the behavioral context 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?

Purpose and the sibling exclusion are front-loaded, and the dense paragraph is largely information-bearing rather than filler. It is longer than strictly necessary (the full field enumeration of the return could be trimmed), so it falls just short of ideal.

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?

With no output schema, the description compensates by spelling out the returned fields (id, status, title, body, reviewers, importance, metadata, expires_at, created_at) and the required poll step. Auth scope, error codes, and limits are all covered, so nothing needed 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%, so the baseline is 3, but the description adds real semantics beyond the schema: agent keys may name only their configured reviewers (403 otherwise) and the reviewer pool is 1-10 external ids with first-decision-wins. Those constraints meaningfully sharpen how reviewers should be supplied.

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 (ask to approve/reject) and resource (an action about to be taken outside Cue), with concrete examples (deleting data, spending money, contacting many people). It also explicitly distinguishes itself from the informing siblings by naming send_event and send_message.

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 exclusion ('Not for informing anyone: use send_event or send_message') plus the follow-up workflow ('poll get_approval with the id and act only on approved'). When-to-use, when-not-to-use, and the alternative tools are all named.

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