Use chirrop-skills

chirrop-skills is the recommended way to make ChirrOp feel native inside OpenClaw.

It gives OpenClaw an opinionated workflow for:

canonical event names
analytics comparisons
valid policy patterns
destination setup and testing
alert triage and delivery inspection
public dashboard page handoff

Install the skills


# Install all skills
npx skills @chirropjs/chirrop-skills
 
# Or install individual skills
npx skills @chirropjs/chirrop-skills/chirrop
npx skills @chirropjs/chirrop-skills/chirrop-monitoring
npx skills @chirropjs/chirrop-skills/chirrop-destinations
npx skills @chirropjs/chirrop-skills/chirrop-alert-triage

Recommended starting path

export CHIRROP_API_KEY
start with the chirrop skill
emit events first and confirm the event definition appears
use the matching ChirrOp SDK for the runtime when possible
run analytics comparisons when the task is periodic comparison rather than thresholding
create policies and destinations when threshold-based alerting matters
direct humans to the public dashboard page when they need to inspect trends

Which skill to use

chirrop: default starting point for logging, reading event definitions, running first-pass analytics comparisons, and creating first policies
chirrop-monitoring: best for recurring comparisons, dashboards, public dashboard pages, market monitoring, sentiment workflows, tool failures, and latency tracking
chirrop-destinations: focused on Slack, Discord, Telegram, email, and generic destination setup
chirrop-alert-triage: focused on reading active alerts, checking deliveries, acknowledging, resolving, and archiving

OpenClaw task mapping

OpenClaw task	Skill
log events, confirm event definition, first analytics check, first policy	`chirrop`
recurring comparisons, charts, public dashboard page handoff	`chirrop-monitoring`
create or test destinations	`chirrop-destinations`
investigate an already-triggered alert	`chirrop-alert-triage`

What to tell OpenClaw

These are good prompts or task instructions for an OpenClaw workflow that has chirrop-skills available:

“Send OpenClaw task duration and error events to ChirrOp”
“Compare the last hour of OpenClaw tool errors with the previous hour using ChirrOp analytics”
“Create a policy for OpenClaw tool failures in the last hour”
“Set up a Slack destination for ChirrOp alerts”
“Show me the public dashboard page for OpenClaw task latency trends”

Why this helps

Without skills, each OpenClaw workflow may invent its own event names and policy patterns.

With chirrop-skills, OpenClaw gets consistent guidance:

event names are the schema
the stable event identity is agent + event
use canonical OpenClaw events first
do not encode time windows into event names
use rollups and policies for hourly or daily logic
use analytics previous windows for explicit comparison queries
treat meta as inspection context in v1

The skills also now align with the ChirrOp SDKs for:

event lookup by stable agent + event
analytics queries
policy creation
destination tests returning a synthetic test alert_id

Contract source of truth

chirrop-skills should follow the contract defined in the docs site.

use docs as the normative source for event identity, periods, analytics window conventions, destination terminology, public dashboard pages, and delivery-history kinds
treat chirrop-skills as the operational layer OpenClaw follows when applying that contract

For testing, start with guided mode first:

explicitly tell OpenClaw to use chirrop-skills
verify it emits canonical event names and valid policies
then repeat in natural mode and confirm it still follows the same contract

Recommended docs flow

For the best OpenClaw integration path, read these pages in order: