chirpier-skills

chirpier-skills is the agent-oriented operational guide for using Chirpier correctly.

It is especially useful for OpenClaw, but the model applies to other agent systems too.

Use chirpier-skills when you want to

• keep event names canonical across agent tasks
• create policies that match the raw stream shape
• validate destinations before relying on real alerts
• triage delivery history correctly
• keep agent prompts aligned with the actual Chirpier API contract

What chirpier-skills is

Think of chirpier-skills as the workflow layer on top of Chirpier’s APIs and SDKs.

• The API Reference defines the HTTP contract.
• SDKs define the language clients.
• chirpier-skills teaches agents how to use those surfaces correctly.

What it enforces

• event names are the schema
• policies should use aggregates that match the raw stream shape
• analytics comparisons should use previous, not invented event names
• destinations should always be tested before production rollout
• alert triage should interpret delivery history with kind=alert, kind=test, and kind=all correctly

Current terminology

Use the current Chirpier product language:

• policy, not monitor
• destination, not connector
• event logs or rollups, not ad hoc time-window streams

Older material may still mention monitor or connector, but the current docs and API use policy and destination.

Read in this order

1. Event Taxonomy
2. Policy Patterns
3. Destinations and Triage
4. OpenClaw Guides

Best fit with OpenClaw

Use OpenClaw Guides when you want the concrete installation path, guided validation flow, and workload-specific playbooks.

Use this page and the rest of Agentic Workflows when you want the reusable Chirpier model first.