Amplitude Analytics connector for AI agents

Tools your analytics agent reaches for on Amplitude Analytics, scoped per tenant.

CALL ANY TOOL

Product analytics on the REST APIs: event segmentation, funnels, retention, cohorts, taxonomy, chart annotations, session replay, and exports. Keys stay vaulted.

amplitudeanalytics_bulk_assign_annotation_category

Bulk assign annotation category

Assign an existing annotation category to multiple annotations at once.

Parameters

Name

Type

Required

Description

annotation_ids

array

Required

Array of annotation IDs to assign to this category.

category_id

integer

Required

The ID of the category to assign to the given annotations.

amplitudeanalytics_create_annotation

Create annotation

amplitudeanalytics_create_dsar_request

Create dsar request

amplitudeanalytics_create_event_property

Create event property

amplitudeanalytics_create_group_property

Create group property

Create a new group property in Amplitude's Taxonomy — a custom property scoped to a specific group type (e.g. 'org', 'company') rather than to users or events. Only group_property and group_type are explicitly confirmed by Amplitude's group-property docs; the remaining descriptive fields (description, type, regex, enum_values, is_array_type, is_hidden, classifications) are inferred by analogy with the sibling event/user property taxonomy endpoints and are not explicitly documented for group properties — an incorrect one is simply rejected with a 400, not destructive. No delete or restore endpoint exists for group properties.

Parameters

Name

Type

Required

Description

group_property

string

Required

The name of the group property to create. Per Amplitude's Taxonomy docs, custom group properties may need a 'grp:' prefix (platform-defined properties do not) — this field is confirmed by Amplitude's docs, but verify the exact prefix convention against your account.

classifications

string

Optional

Data classification tags to apply to this property (e.g. PII markers), matching whatever classifications are configured in the Amplitude project's data governance settings. INFERRED from the sibling event/user property taxonomy endpoints — Amplitude's group-property docs do not confirm this field exists for group properties; verify empirically before relying on it.

description

string

Optional

Human-readable description of what this property represents. INFERRED from the sibling event/user property taxonomy endpoints — not explicitly confirmed for group properties by Amplitude's docs; verify empirically before relying on it.

enum_values

string

Optional

Allowed values for this property when type is 'enum'. INFERRED from the sibling event/user property taxonomy endpoints, including the assumption that it's a plain string here rather than an array — not explicitly confirmed for group properties by Amplitude's docs; verify the exact expected format empirically.

group_type

string

Optional

The group type this property belongs to (e.g. 'org', 'company'). Confirmed by Amplitude's Taxonomy docs as a valid create field. Returns 404 if the group type doesn't exist, 409 if the property already exists on that type.

is_array_type

boolean

Optional

Whether this property holds an array of values rather than a single value. INFERRED from the sibling event/user property taxonomy endpoints — not explicitly confirmed for group properties by Amplitude's docs; verify empirically before relying on it.

is_hidden

boolean

Optional

Whether this property is hidden from Amplitude's UI pickers (e.g. the chart builder's property list). INFERRED from the sibling event/user property taxonomy endpoints — not explicitly confirmed for group properties by Amplitude's docs; verify empirically before relying on it.

regex

string

Optional

A regular expression that values of this property must match. INFERRED from the sibling event/user property taxonomy endpoints — not explicitly confirmed for group properties by Amplitude's docs; verify empirically before relying on it.

type

string

Optional

The data type of this property. INFERRED from the sibling event/user property taxonomy endpoints — not explicitly confirmed for group properties by Amplitude's docs; verify empirically before relying on it.

amplitudeanalytics_delete_annotation

Delete annotation

amplitudeanalytics_delete_event_category

Delete event category

amplitudeanalytics_delete_event_type

Delete event type

amplitudeanalytics_get_active_and_new_user_counts

Get active and new user counts

amplitudeanalytics_get_annotation_category

Get annotation category

amplitudeanalytics_get_chart_results

Get chart results

amplitudeanalytics_get_cohort_usage

Get cohort usage

amplitudeanalytics_get_event_category

Get event category

amplitudeanalytics_get_event_streaming_metrics

Get event streaming metrics

amplitudeanalytics_get_funnel_results

Get funnel results

amplitudeanalytics_get_retention_analysis

Get retention analysis

Pull the Retention Analysis chart from the Amplitude Dashboard REST API: what fraction of users who did a 'start' action came back to do a 'return' action, over a date range, with optional bracket/rolling/n-day retention modes, segment filters, and one group-by property. Response shape: {"data": {"series": [{"dates": [string,...], "values": {date: [{"count": number, "outof": number, "incomplete": boolean}]}, "combined": [{"count": number, "outof": number, "incomplete": boolean}]}], "seriesMeta": [{"segmentIndex": number, "eventIndex": number}]}}. Rate limits: 5 concurrent requests shared with other Amplitude Dashboard/Cohort API calls, up to 108,000 cost per hour.

Parameters

Name

Type

Required

Description

end

string

Required

End date of the query range, format YYYYMMDD (e.g. 20260131).

return_event

string

Required

JSON-encoded event object for the 'returning' action. Amplitude documents two special event_type values here: '_all' (all events) or '_active' (all active events). A regular event definition like {"event_type":"opened_app"} is also accepted.

start

string

Required

Start date of the query range, format YYYYMMDD (e.g. 20260101).

start_event

string

Required

JSON-encoded event object for the 'start' action. Amplitude documents two special event_type values here: '_new' (users new in the date range) or '_active' (all active users). A regular event definition like {"event_type":"sign_up"} is also accepted.

group_by

string

Optional

Property to group results by, e.g. gp:country. Limited to one property.

interval

integer

Optional

Time bucket size: 1 (daily), 7 (weekly), or 30 (monthly). Defaults to 1.

retention_brackets

string

Optional

JSON-encoded array of day brackets, required only when retention_mode is 'bracket'. Format: [[0,4]] for a single 0-4 day bracket; multiple brackets can be listed.

retention_mode

string

Optional

Retention calculation type: bracket (returning within specific day brackets, requires retention_brackets), rolling (returning at any point after, unbounded), or n-day (returning on exactly day N). Defaults to n-day.

segment_definitions

string

Optional

JSON-encoded array of segment definitions to split retention by cohort/property segments, e.g. [{"prop":"gp:country","op":"is","values":["US"]}].

amplitudeanalytics_get_session_replay_files

Get session replay files

amplitudeanalytics_get_user_mapping

Get user mapping

amplitudeanalytics_list_annotation_categories

List annotation categories

amplitudeanalytics_list_cohorts

List cohorts

amplitudeanalytics_list_event_types

List event types

amplitudeanalytics_list_user_deletion_jobs

List user deletion jobs

amplitudeanalytics_remove_user_from_deletion

Remove user from deletion

amplitudeanalytics_request_cohort_membership

Request cohort membership

amplitudeanalytics_restore_event_type

Restore event type

amplitudeanalytics_search_users

Search users

amplitudeanalytics_update_annotation

Update annotation

amplitudeanalytics_update_cohort_membership

Update cohort membership

amplitudeanalytics_update_event_type

Update event type

Partially update an existing event type in Amplitude's taxonomy. Only the fields you provide are changed; omitted fields keep their current value. Set new_event_type to rename the event type.

Parameters

Name

Type

Required

Description

event_type

string

Required

The name of the existing event type to update.

tags

string

Optional

New comma-separated list of tags for this event type.

amplitudeanalytics_upload_cohort

Upload cohort

Create a new Amplitude behavioral cohort from an explicit list of user or Amplitude IDs, or update an existing cohort's membership list wholesale by passing existing_cohort_id. To add/remove individual members from an already-created cohort instead, use amplitudeanalytics_update_cohort_membership. CONFIRMED (live-tested against a real project, App ID and real non-dry-run save attempt, not just skip_save): every id in ids is validated against Amplitude's actual known-users table before any save happens — IDs with zero tracked activity in the project are rejected with error code 120 ("User id is invalid") or 110 ("Amplitude id is invalid"), reporting {matched, totals, invalid_ids_sample}. This rejection happens even with skip_invalid_ids: true when 100% of the provided IDs are invalid (there's nothing left to keep) — skip_invalid_ids only helps when the ids list is a genuine mix of valid and invalid entries. There is no way to create a cohort from IDs that have no real tracked history in the project.

Parameters

Name

Type

Required

Description

app_id

integer

Required

The Amplitude project id that will contain the cohort.

id_type

string

Required

The type of identifier used in the ids list: BY_AMP_ID (Amplitude IDs) or BY_USER_ID (your own user IDs).

ids

array

Required

One or more user or Amplitude IDs to include in the cohort, matching the type set in id_type. CONFIRMED: each ID must already have real tracked activity in the Amplitude project — Amplitude validates against its known-users table and rejects (error 120/110) any ID it has never seen, even a syntactically valid-looking one.

name

string

Required

Display name for the new cohort.

owner

string

Required

published

boolean

Required

If true, the cohort is discoverable by other users in the Amplitude project. If false, it's hidden.

cohort_group

string

Optional

Name of an existing group to base the cohort on, if creating a group-based cohort rather than a user-based one.

existing_cohort_id

string

Optional

If set, replaces this existing cohort's membership list instead of creating a new cohort.

skip_invalid_ids

boolean

Optional

If true, invalid IDs are silently skipped and the cohort is created from the remaining valid ones. If false, the entire request is rejected if any ID is invalid. Defaults to true. CONFIRMED: this only helps with a genuine mix of valid and invalid IDs — if every ID in the list is invalid (0 matched), the request is rejected regardless of this setting, since there would be nothing left to build the cohort from.

skip_save

boolean

Optional

If true, validates the request without actually saving the cohort — useful for a dry run. Defaults to false.

For more tools, view docs.

Build your Agent

Drop the toolkit in, point it at the tenant, and your agent can run segmentation and funnel queries from the first run.

import { ScalekitClient } from "@scalekit-sdk/node";
import { DynamicStructuredTool } from "@langchain/core/tools";
import { createReactAgent } from "@langchain/langgraph/prebuilt";
import { z } from "zod";

const sk = new ScalekitClient(envUrl, clientId, clientSecret);

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["amplitudeanalytics"], toolNames: ["amplitudeanalytics_get_event_segmentation", "amplitudeanalytics_get_funnel_results", "amplitudeanalytics_list_cohorts"] },
  pageSize: 100,
});

const lcTools = tools.map((t) => new DynamicStructuredTool({
  name: t.tool.definition.name,
  description: t.tool.definition.description,
  schema: z.object({}).passthrough(),
  func: async (args) => {
    const { data } = await sk.tools.executeTool({
      toolName: t.tool.definition.name,
      identifier: "user_123",
      params: args,
    });
    return JSON.stringify(data);
  },
}));

const agent = createReactAgent({ llm, tools: lcTools });

import { ScalekitClient } from "@scalekit-sdk/node";
import OpenAI from "openai";

const sk = new ScalekitClient(envUrl, clientId, clientSecret);
const openai = new OpenAI();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["amplitudeanalytics"], toolNames: ["amplitudeanalytics_get_event_segmentation", "amplitudeanalytics_get_funnel_results", "amplitudeanalytics_list_cohorts"] },
  pageSize: 100,
});

const llmTools = tools.map((t) => ({
  type: "function",
  function: {
    name: t.tool.definition.name,
    description: t.tool.definition.description,
    parameters: t.tool.definition.input_schema,
  },
}));

const resp = await openai.responses.create({
  model: "gpt-4o", input: prompt, tools: llmTools,
});

import { ScalekitClient } from "@scalekit-sdk/node";
import Anthropic from "@anthropic-ai/sdk";

const sk = new ScalekitClient(envUrl, clientId, clientSecret);
const anthropic = new Anthropic();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["amplitudeanalytics"], toolNames: ["amplitudeanalytics_get_event_segmentation", "amplitudeanalytics_get_funnel_results", "amplitudeanalytics_list_cohorts"] },
  pageSize: 100,
});

const llmTools = tools.map((t) => ({
  name: t.tool.definition.name,
  description: t.tool.definition.description,
  input_schema: t.tool.definition.input_schema,
}));

const msg = await anthropic.messages.create({
  model: "claude-sonnet-4-6", max_tokens: 1024,
  tools: llmTools,
  messages: [{ role: "user", content: prompt }],
});

import { Agent } from "@google/adk/agents";
import {
  MCPToolset, StreamableHTTPConnectionParams,
} from "@google/adk/tools/mcp";

const toolset = new MCPToolset({
  connectionParams: new StreamableHTTPConnectionParams({
    url: "https://mcp.scalekit.com/amplitudeanalytics",
    headers: { Authorization: `Bearer ${userScopedToken}` },
  }),
});

const agent = new Agent({
  name: "agent", model: "gemini-2.0-flash",
  tools: await toolset.getTools(),
});

Try these prompts

Paste any prompt into your agent to start querying Amplitude from your workflows.

Funnels and retention

Copy the prompt

Copied

Where do users drop off in the onboarding funnel over the last 30 days?

Copy the prompt

Copied

Compare 7-day retention for users who connected a data source against those who did not.

Copy the prompt

Copied

Show conversion for [event_a] to [event_b] split by plan.

Segmentation

Copy the prompt

Copied

Break [event] counts down by country for last week.

Copy the prompt

Copied

Which events grew fastest month over month?

Copy the prompt

Copied

List the top property values sent with [event].

Taxonomy and cohorts

Copy the prompt

Copied

List cohorts in the project with size and last computed time.

Copy the prompt

Copied

Show event types with no schema description in the taxonomy.

Copy the prompt

Copied

Annotate today's chart with the v2.4 release.

SEE HOW AUTH WORKS

Your users connect once. Their Amplitude Analytics credentials stay vaulted, every call is scope-checked, and every action is logged.

Authorize

Your user connects

Amplitude Analytics

once. We tie it to their identity and the meetings they approved — no shared bot account, no org-wide access

Who:

user ‘A’

when:

Once per user

access:

Limited to user

Store

Their

Amplitude Analytics

token lives in a vault scoped to them. User A's meetings are never reachable by an agent acting for user B, even on the same connection

vault:

encrypted

scope:

per-user

tokens:

auto-refreshed

Resolve

When your agent calls a

Amplitude Analytics

tool, we fetch the right token server-side. It never touches your agent, never appears in the LLM context, never shows up in your logs

speed:

~40ms

check:

before every call

seen by:

nobody

Audit

Every

Amplitude Analytics

tool call is logged — who triggered it, which meeting was fetched, what came back. 90 days of history, tied to the user who authorized it

history:

90 days

export:

SIEM-ready

logged:

every call

Test other agents

See the same per-user auth pattern across other connectors.

GTM and RevOps Teams

Revenue forecast commentary

Pulls open pipeline from Salesforce and HubSpot, calculates coverage against quota, flags at-risk stages, posts commentary to Slack, and logs every snapshot to Google Sheets.

GTM and RevOps Teams

Deal intelligence agent

Pulls recent Gong calls, scores deal risk with an LLM, cross-references the record in Attio, and DMs each owner their at-risk deals in Slack. Every read is scoped to that rep's own access.

Engineering Teams

Engineering standup agent

Pulls commits from GitHub and GitLab, tracks issue movement in Jira, and posts a per-engineer standup brief to Slack. Each engineer's activity is read on their own delegated OAuth.

GTM and RevOps Teams

Competitive intelligence briefing agent

Scans Gong calls for competitor mentions, matches each one to its Notion battlecard, and DMs every affected rep a single Slack digest per cycle. Every call runs as the PMM who owns the briefing, never a shared bot.

Why Scalekit

Secure your agent's access. Connectors ship in minutes

01.

Shared tokens break per-user analytics

A shared token looks fine in a demo. In production every call looks like a service account. Scalekit resolves the real user credential so attribution, audit, and scope stay accurate.

// shared token
audit → bot_service_account
user_filter → broken

// scalekit
audit → user_abc
scope → enforced ✓

02.

Authentication is not authorization

03.

Multi-tenancy is architectural

04.

Amplitude Analytics today. Others tomorrow.

“Our agents act across Salesforce, Gong, Google Drive, and more, on behalf of every customer. Scalekit behind the scenes meant we can keep adding tools without ever rebuilding how credentials or tool calling work.”

Venu Madhav Kattagoni

Head of Engineering / Von

FAQs

Frequently Asked Questions

Does the agent access Amplitude Analytics as the user or as a shared key?

As the user. Each workspace member authorizes once and Scalekit resolves their credential at request time. Audit logs attribute every action to that user, not a shared service account.

Where is the Amplitude Analytics API key stored?

In Scalekit's managed AES-256 token vault, namespaced per tenant. Refresh is automatic. Revocation is a single dashboard action. Tokens never appear in prompts, logs, or LLM context.

Can I limit what the agent is allowed to do in Amplitude Analytics?

Yes. Pass a tool name filter to listScopedTools so the analytics agent only sees the subset you authorize. Pre-API-call scope checks block out-of-policy actions before the request reaches Amplitude Analytics.

What happens when a user revokes Amplitude Analytics access?

The connection is invalidated on the next tool call. Subsequent requests for that user fail closed with a clear error. Other users in the tenant remain unaffected. The event is logged for audit.

Do analytics reads stay scoped to one Amplitude project?

Yes. The key pair you register is project scoped, and Scalekit stores it in that tenant's vault namespace. An agent call for one tenant resolves only that tenant's credential, so queries can never read another project's events.

On this page

Tool and actions Framework Compatibility Starter Prompt Auth Lifecycle Try Other Agents Why Scalekit