AI Agent Ready
Structured JSON outputs designed for AI agents. Works with OpenClaw, Claude Code, and any LLM-based agent.
AI Agents CLI
Marketing data CLI
for autonomous AI agents
Query and manage 40+ marketing platforms from your terminal. Built for OpenClaw, Claude Code, and any shell-capable AI agent.
$ npm install -g insightfulpipeGet Started In Seconds
Install, authenticate, and start querying
The CLI is designed for both humans and agents: simple commands, predictable outputs, and a workflow that reveals the schema before execution.
terminal
# Install the CLI
$ npm install -g insightfulpipe
# Authenticate with your API token
$ insightfulpipe login
? Enter your API token (ip_sk_...): ********
✓ Authenticated as My Agency
# List connected platforms
$ insightfulpipe platforms
google-analytics Google Analytics 4 3 properties
facebook-ads Facebook Ads 2 ad accounts
shopify Shopify 1 store
# Query your data
$ insightfulpipe query google-analytics \
-b '{"action":"run_report","date_range":"last_30_days"}'
{ "sessions": 12847, "users": 9231, "bounce_rate": 0.42 }
Supported Platforms
40+ data sources
Pull analytics, advertising, ecommerce, CRM, publishing, and utility data through one consistent command interface.
Google AnalyticsGoogle AdsGoogle Search ConsoleGoogle SheetsGoogle My BusinessFacebook AdsFacebook PagesInstagramTikTok AdsTikTok PagesLinkedIn AdsLinkedIn PagesYouTubePinterest AdsSnapchat AdsMicrosoft AdsX AdsShopifyWooCommerceMagentoHubSpotKlaviyoMailchimpStripePayPalSlackTelegramWordPressCallRailDataForSEOApple App StoreGoogle PlayPageSpeed InsightsWeb CrawlerAds LibrariesEnrichment Crawlers
Built for Automation
Designed around how agents reason
The CLI exposes account discovery, schema inspection, and execution as separate steps so agent loops stay predictable and debuggable.
40+ Platforms
Query Google Analytics, Facebook Ads, Shopify, HubSpot, and dozens more from a single CLI.
Secure by Default
API tokens stored securely in your system keychain. No credentials in environment variables.
Discover-Schema-Execute
Discover accounts, inspect schemas, then query. It matches how autonomous agents reason through work.
Read and Write Operations
Pull reports, create posts, update records, and trigger actions across supported platforms.
OpenClaw Skill
Published on ClawHub as a skill, so OpenClaw-compatible agents can discover and use it automatically.
Workflow
Discover, inspect, then execute
AI agents do better when they can inspect structure before taking action. The CLI is built around that exact loop.
1
Discover accounts
List connected platforms and identify the workspace, brand, and account IDs an agent needs before running queries.
$ insightfulpipe accounts \
--platform google-analytics
# Returns workspace, brand, and account IDs
2
Inspect schema
Pull the exact request structure, supported actions, required fields, and available metrics before an agent decides what to send.
$ insightfulpipe helper \
google-analytics
# Returns available actions and fields
3
Query or execute
Run read queries or write actions
with the schema the agent
already validated.
$ insightfulpipe query \
google-analytics -b '{...}'
# Returns structured JSON data
OpenClaw
Published as a skill
InsightfulPipe is published on ClawHub so OpenClaw-compatible agents can discover it automatically instead of relying on hardcoded integrations.
# Install InsightfulPipe as an OpenClaw skill
$ npm install -g insightfulpipe
# Set your API token
$ export INSIGHTFULPIPE_TOKEN=ip_sk_...
# Your agent can now discover and use InsightfulPipe
$ insightfulpipe doctor
✓ Token valid
✓ 12 platforms connected
✓ SKILL.md found — agent-ready
Claude Code
Also works with Claude Code
Claude Code can execute the CLI directly to pull reports,
compare sources, and take follow-up actions from your terminal workflow.
# Ask Claude to analyze your marketing data
> Pull last month's Google Analytics report
> and compare it with Facebook Ads spend
# Claude autonomously runs:
$ insightfulpipe query google-analytics \
-b '{"action":"run_report","date_range":"last_month"}'
$ insightfulpipe query facebook-ads \
-b '{"action":"campaigns","date_range":"last_month"}'
✓ Reports fetched and compared. Here's the analysis...
Ready to try it?
Install the CLI and let your agent work from real marketing data.
Start with the npm package, wire in your token, and give your terminal-based agents direct access to connected platforms.