axiom_skill

Overview

This skill integrates with Axiom, a cloud-native observability platform, to ingest, query, and manage logs and event data via the REST API. It provides command-line-ready examples for dataset management, batch ingestion (JSON/NDJSON), APL queries, annotations, and monitors. Use it to automate observability pipelines and programmatic analytics.

How this skill works

The skill issues REST calls to Axiom API endpoints and edge ingest endpoints for high-throughput writes. It demonstrates listing and creating datasets, ingesting JSON or NDJSON payloads, running APL (Axiom Processing Language) queries including aggregations, and managing annotations and monitors. Examples use bash -c wrappers so environment variables remain available when piping to curl.

When to use it

• When you need programmatic ingestion of logs, metrics, or events into Axiom
• When you want to run ad hoc or automated APL queries for analysis and dashboards
• When creating or managing datasets, monitors, or annotations via scripts
• When building CI/CD or observability pipelines that require automated telemetry delivery
• When you want to batch events for performance-sensitive ingestion

Best practices

• Use regional edge ingest endpoints (us-east-1 or eu-central-1) for writes; reserve api.axiom.co for control plane requests
• Batch multiple events per request (JSON array or NDJSON) to improve throughput and reduce overhead
• Always include explicit timestamps in events; otherwise server receive time will be applied
• Set and monitor X-RateLimit-Remaining to avoid throttling and implement exponential backoff on 429 responses
• Specify startTime and endTime in APL requests to reduce query scope and improve performance

Example use cases

• Automated log collection from servers and containers into a named Axiom dataset for centralized analysis
• Nightly aggregation jobs that use APL to compute counts, averages, or hourly summaries for dashboards
• Deployment annotations written at release time to correlate incidents with deploy events
• Scripting dataset lifecycle: create, inspect, and delete datasets as part of test or staging workflows
• Alerting pipelines that list monitors and create annotations when threshold conditions are met