agentsclimarketplace

Log aggregation data simulation

Skill kjuhwa/skills-hub/skills/workflow/log-aggregation-data-simulation

Self-correcting knowledge corpus for Claude Code — 9 stable shape clusters, bias-correction pipeline baked into contribution flow. 47 papers, 45 techniques, 1.1k skills.

Install
npx -y skills add kjuhwa/skills-hub --skill log-aggregation-data-simulation

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generating realistic synthetic log streams with bursts, level distributions, and source correlation for demo UIs

SKILL.md

1.8 KB, as published. Nobody here has run it

log-aggregation-data-simulation

Realistic log aggregation demos require simulation that matches production statistical shape, not uniform random noise. Generate logs with a Poisson base rate per source (e.g., 5-50 logs/sec/service) multiplied by a diurnal sine envelope (higher daytime volume), then inject burst events: correlated error storms where one upstream service failure triggers elevated error rates in 2-4 downstream services within a 10-60 second window. Level distribution should follow roughly info 70%, debug 15%, warn 10%, error 4%, fatal 1% during normal operation, shifting to error 30-50% during burst windows. Without bursts and correlation, the heatmap looks flat and the stream river looks like static — operators instantly recognize this as fake.

Log message bodies should draw from a templated pool per service ("GET /api/users/{id} 200 {ms}ms", "Connection pool exhausted: {n}/{max}", "JWT expired for user={uid}") with a small set of structured fields (trace_id, span_id, user_id, host) so query-console filters have something meaningful to match. Reuse trace_ids across 3-8 correlated log entries to demonstrate distributed tracing joins. Seed the RNG so demos are reproducible, but expose a "live mode" that streams new entries on a wall-clock interval (50-200ms ticks) with a rolling window (last 10k-100k entries) to avoid unbounded memory. Pre-generate 1-2 hours of historical data on page load so the heatmap has content before the live stream fills it in.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.