Mingri DMP Audience Targeter for Openclaw

An AI agent skill that integrates with the Mingri DMP API to build, expand, and map precise target audiences using six different advanced selection rules.

mingri26
v1.0.0
Jun 12, 2026
0
477
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install dmp-segment

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install dmp-segment using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Mingri DMP Audience Targeter?

The Mingri DMP Audience Selection skill is an enterprise-grade digital marketing solution built for the Openclaw Skills ecosystem. By interfacing directly with the Mingri DMP Open Platform API, this skill empowers marketing and analytics agents to bypass manual visual interfaces and generate hyper-targeted customer segments directly through conversational inputs.

The skill supports six specialized audience selection pipelines: behavioral tag combinations, ad engagement tracking (impressions and clicks), LBS geographical perimeters, mobile app installation matrices, lookalike audience matching, and TV-to-Mobile cross-screen mapping. Incorporating a progressive design philosophy, the skill automatically resolves dependency configurations, checks parameters, and ensures structural logical consistency prior to hitting remote APIs.

Mingri DMP Audience Targeter Use Cases

  • Creating custom multi-conditional audience segment combinations utilizing and/not demographics and tag interests.
  • Re-targeting users who engaged with specific campaign or spot IDs within custom date frames (up to 62 days).
  • Discovering competitive insights and capturing target market share by querying app install statuses across similar devices.
  • Setting up location-based circular or polygonal perimeters to capture regional offline store footprints (LBS targeting).
  • Executing lookalike calculations on existing seed populations using machine-learning constraints for campaign scaling.
  • Mapping active OTT/Smart TV hardware targets directly back to mobile devices for cohesive multi-screen digital experiences.

How Mingri DMP Audience Targeter Works

  1. Intent Classification: The skill dynamically detects user requests, identifying whether the query targets skill definitions, segment status retrieval, or raw audience creation pipelines.
  2. Authentication Audit: Scans local directory environments to confirm authorization plugins are installed and evaluates credential statuses.
  3. Rule Configuration Mapping: Determines the target audience model matching your strategy requirements and steps you through platform parameters.
  4. Fallback Handling: Gracefully handles API retrieval failures during dictionary requests, presenting downgrade strategies rather than halting agent executions.
  5. Structural Validation Check: Assesses numeric parameters, dates, region dimensions, and formatting structures to ensure perfect API compatibility.
  6. Double-Confirmation Display: Builds a clear summary matrix of the target payload configuration, pausing execution to await explicit user approval.
  7. Task Execution: Invokes script clients to execute safe HTTP requests to the Mingri DMP Platform and returns the pending task audienceId.
  8. Transaction Logging: Hooks into local logging skills to preserve session configurations and appends calculation tracking markers locally.

Mingri DMP Audience Targeter Setup

Prerequisites

Prior to using this skill, secure an API Access Key and Secret Key from the Mingri Data support team. Apply using one of the following methods:

Installation

Set up the audience engine alongside its primary secure auth module and history logging dependencies. Run the following commands within your CLI tool environment:

# Install the main audience selection script catalog
openclaw skills install mingri-dmp-audience-targeter

# Install the authentication agent (Required dependency)
openclaw skills install mingdata-dmp-auth

# Install the task logging agent (Highly recommended dependency)
openclaw skills install skill-logger

Access Configuration

On initial activation, the underlying Openclaw Skills core will prompt you to input your authorized credentials. These keys are stored locally and kept secure at ~/.mingdata_dmp_credentials:

Access Key: AK-xxxxxxxxxxxxxxxx
Secret Key: SK-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Mingri DMP Audience Targeter Data Schema & Taxonomy

API Data Schema & Payload Requirements

Each audience selection type maps directly to a specified analytical JSON target payload. Below are the key structure schemas handled automatically by the agent.

APP Rule Audience Data Segment (APP Dimension)

Parameter Type Required Description Example / Values
name String Yes Unique naming descriptor of the segment. "ecommerce_active_users"
trackType String Yes Underlying device tracker target. "MOBILE", "PC", "OTT"
idTypes Array Yes Target tracking identity hash lists. ["MD5_IMEI", "MD5_OAID"]
data.dimension String Yes Segment dimension constraint key. Must be set to "APP"
data.type Integer Yes Underlying targeting analytical criteria. 0 (Active user), 1 (Installed app)
data.packageNames Array Yes Listing of distinct targeted application package IDs. ["com.taobao.taobao", "com.tmall.wireless"] (Min 4 package IDs)

Geolocation Rule Audience Segment (LBS Dimension)

Parameter Type Required Description Constraints
name String Yes Name of segment Max 50 characters
drawMode String Yes Geo boundary selection methodology. "batch" (circle), "manual" (polygon)
dataOrigin String Yes Tracking target location origin parameter. "recentArrive", "homeAddress", "workAddress"
radius Integer Yes Circle expansion perimeter scope (meters). Integer value between 150 and 6000
timeRange String Yes Custom hourly range boundaries. Format: "YYYY-MM-DD HH"

Local Log Registry Format

Task details logged via the logging skill are kept under ~/.skill-logger/task_history.json configured in this format:

{
  "taskId": "audienceId_20260529090454",
  "name": "Targeted Lookalike Segment",
  "createdTime": "2026-05-29 09:04:54",
  "configuration": {
    "expandType": 0,
    "confidenceMin": 0.5,
    "confidenceMax": 0.9
  }
}

Mingri DMP Audience Targeter Advanced Features

  • Dynamic Dimension Sourcing: Maps localized geographic names and coordinates to proprietary DMP regionId dictionaries, maintaining precise lookups with static fallback structures.
  • Degraded Execution Workflows: If remote dictionary requests fail mid-session, the agent suggests recovery solutions like property exclusions or customized manual parameter overrides to ensure process continuity.
  • Tabular Pre-flight Summarization: Compiles parameters into a structural display table, requesting formal approval to proceed before invoking the remote API.
  • Downhill Strategy Activation: Recommends related Openclaw Skills tools to execute adjacent analytics or campaign delivery tasks as soon as segment generation completes.
  • Secure Auth Compartmentalization: Segregates sensitive key storage strictly within the localized config client (mingdata-dmp-auth) to keep credentials protected.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*