Nephia

Change keyword settings

keyword_update

DestructiveIdempotent

Change a keyword's settings: name, default interval, sorting, sentiment, co-brands, agent step, subject role, mute rules, VIP authors, rules, webhook, schedule and channels. A key you leave out is left alone; null clears it. Never its Sources or searches: that is keyword_source_set, so this tool starts no search over. Call keyword_estimate with the same arguments first: it runs the same validation, writes nothing, and returns the price and the estimateToken this tool requires. Starts or changes polling that is charged per check at each Source's credit rate (AI answers per engine asked) until the keyword or the Source is paused. Call keyword_estimate first, show the person the monthlyCredits it returns, and send its estimateToken only after they agree.

Input Schema

TableJSON Schema

Name	Required	Description
`id`	Yes	Keyword id, from list_keywords. Another account's id is a 404, never a 403.
`name`	No
`color`	No	The keyword colour in the dashboard. Left out on create, an unused one is picked.
`rules`	No	Replaces the list. Copy the shape from get_keyword on an existing keyword.
`aiStep`	No	Your own question asked of every mention, with the fields to fill: {"instruction": one or two sentences, 800 characters at most, "schema": 1 to 12 flat fields, each {"type": "string" \| "number" \| "boolean" \| "enum", "description", "values" for an enum, "maxLength" for a string}}, no nesting and no arrays. The answer arrives on each event as analysis. Costs 1 credit per 25 items, on top of the check rate. Delivery waits up to 120s for it; on timeout the event is still delivered, with the answer null and a status saying why. null removes it.
`aiPreset`	No
`aiPrompt`	No
`channels`	No	Replaces the delivery channels attached to the keyword.
`schedule`	No	An activation window with ISO instants. null keeps the keyword always on.
`aiEnabled`	No	Sort mentions into buckets.
`muteRules`	No	Replaces the list. A plain string mutes that word.
`vipAuthors`	No	Replaces the list. A plain string is a handle.
`webhookUrl`	No	A publicly reachable HTTPS endpoint the person controls: new mentions are POSTed there. Never invent this URL, ask the person for it. Left out, the keyword records what it catches without pushing it, and you read it with list_mentions, get_keyword_results or get_keyword_events. null removes the webhook.
`subjectRole`	No	What the keyword is about: your own brand, a competitor's, or a topic. A new topic without globalCriteria.match is created with match word.
`webhookMode`	No	all delivers every new mention; rules delivers only what a rule routes.
`estimateToken`	Yes	The estimateToken keyword_estimate returned for exactly these arguments. Valid 15 minutes. A token for other arguments is refused: estimate again.
`coBrandsEnabled`	No	AI answers only: also read each answer for the other brands it names, charged per answer on top of the run.
`relevanceContext`	No	One sentence the AI reading uses to judge relevance, such as "Acme is a cloud storage company, not the cartoon." null clears it.
`sentimentEnabled`	No	Read every mention for sentiment and intent.
`refreshIntervalSeconds`	No	Default seconds between checks, for every Source without its own interval. The plan and each Source set a floor; the estimate names the band when a value is outside it.

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedSep 26, 2026

changedInput schema / properties / subjectRole / description

Previous value: -"What the keyword is about: your own brand, a competitor's, or a topic."New value: +"What the keyword is about: your own brand, a competitor's, or a topic. A new topic without globalCriteria.match is created with match word."

Changed2 schema fields changedSep 24, 2026

addedInput schema / properties / color

Added value: +{
+  "description": "The keyword colour in the dashboard. Left out on create, an unused one is picked.",
+  "enum": [
+    "keyword-1",
+    "keyword-2",
+    "keyword-3",
+    "keyword-4",
+    "keyword-5",
+    "keyword-6",
+    "keyword-7",
+    "keyword-8"
+  ],
+  "type": "string"
+}

addedInput schema / properties / relevanceContext

Added value: +{
+  "anyOf": [
+    {
+      "maxLength": 400,
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "description": "One sentence the AI reading uses to judge relevance, such as \"Acme is a cloud storage company, not the cartoon.\" null clears it."
+}

AddedSep 21, 2026

TDQS

A4.7/5.0

Behavior5/5

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

Annotations declare destructiveHint=true, readOnlyHint=false, idempotentHint=true, and openWorldHint=true. The description adds substantial context beyond these: it reveals that the tool starts or changes polling with per-check credit charges, that estimateToken must exactly match the arguments and expires in 15 minutes, and that null clears fields while omitted keys are untouched. It also confirms it never starts a search. These details align with the annotations and significantly enhance the agent's understanding of side effects and constraints.

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?

The description is long but every sentence carries weight. It is structured logically: purpose, exclusions, workflow, charging model, and a final reminder. It front-loads the primary action and then builds context. No redundant or filler sentences; the density is justified by the tool's complexity.

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?

Given 20 parameters and no output schema, the description covers the critical operational aspects: the mandatory estimateToken prerequisite, the distinction between omit and null, the charging model, and the scope limitation. It does not explicitly describe the return value, but for an update tool this is often a confirmation; the absence is acceptable given the rich workflow guidance. Overall, an agent has enough to invoke it correctly.

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 85%, so most parameters already have descriptive entries. The description adds cross-cutting semantics that apply to all optional fields: 'A key you leave out is left alone; null clears it.' It also clarifies the estimateToken requirement — that it must be for exactly these arguments and is valid 15 minutes. This goes beyond the schema's per-field descriptions and aids correct usage.

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?

The description opens with a clear verb-resource pair — 'Change a keyword's settings' — and enumerates the exact fields (name, interval, sorting, etc.). It explicitly differentiates from keyword_source_set by stating 'Never its Sources or searches: that is keyword_source_set', and distinguishes keyword_estimate as a validation-only call. The purpose is unambiguous and distinct from all sibling tools.

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?

The description gives explicit when-to-use guidance: it instructs to call keyword_estimate first with the same arguments, show the monthlyCredits to the user, and only then send the estimateToken after agreement. It also clearly states what this tool does NOT do (sources/searches belong to keyword_source_set) and explains the omit-vs-null semantics. This leaves no ambiguity about the correct workflow or alternatives.

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

All Nephia tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?