agentsclimarketplace

Actor model data simulation

Skill kjuhwa/skills-hub/skills/workflow/actor-model-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 actor-model-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

Generate realistic actor workloads with crashes, backpressure, and supervision restarts for demos

SKILL.md

2.0 KB, as published. Nobody here has run it

actor-model-data-simulation

Seed simulations with a small cast of archetypal actors rather than random ones: a Producer that emits messages at a configurable Poisson rate, a Worker pool behind a Router actor, a FlakyService that crashes with probability p on each message, and a SlowConsumer whose processing time is sampled from a log-normal distribution. This handful reproduces the interesting phenomena — backpressure, restart storms, head-of-line blocking, supervisor escalation — without needing a large synthetic graph. Parameterize rates, crash probability, and mailbox bound as sliders so users can drive the system into each regime live.

Drive the simulation from a single logical clock with deterministic seeding. Every actor action (send, receive, crash, restart) is scheduled on a priority queue keyed by virtual time, and the RNG is seeded per-run so a given seed reproduces the exact message interleaving. This is what makes actor demos trustworthy — users who pause on a crash can step forward one message at a time and see the supervisor's decision tree (:restart / :stop / :escalate) apply deterministically. Emit events to the same log the visualization consumes so sim and UI share one source of truth.

For the chat-REPL variant, treat each REPL session as an actor with its own mailbox and let users type spawn, send <pid> <msg>, kill <pid>, and tree as commands. Back the REPL with the same simulation engine; the REPL is just another event producer. Pre-canned scripts (demo:restart-storm, demo:backpressure, demo:escalation) should seed the RNG and replay a known scenario so onboarding users hit the interesting failure modes within the first minute.

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.