Small, focused packages for Node.js. Each one does one thing well. Many are distributed, backed by Redis or Postgres with leader election, clustering, and shared state built in. All ESM, zero or minimal dependencies, published under the @prsm scope on npm.
| Package | Test | Version | Description |
|---|---|---|---|
@prsm/ms |
 |
 |
Parse and convert time strings to milliseconds |
@prsm/ids |
 |
 |
Short, obfuscated, collision-proof, reversible identifiers |
@prsm/hash |
 |
 |
Simple salted string hashing on top of node:crypto |
@prsm/from |
 |
 |
Fluent query builder for filtering, transforming, and aggregating arrays of objects |
@prsm/fsm |
 |
 |
Non-deterministic finite state machine with time-based ticks, persistence, and multiple active states |
@prsm/lock |
 |
 |
Distributed locking primitives for Redis - mutex and semaphore |
@prsm/limit |
 |
 |
Redis-backed distributed rate limiting with token bucket, sliding window, and leaky bucket algorithms |
@prsm/cache |
 |
 |
Distributed cache with stampede prevention, stale-while-revalidate, and tag invalidation |
@prsm/trace |
 |
 |
Distributed trace propagation with W3C TraceContext, AsyncLocalStorage, and framework adapters |
@prsm/cron |
 |
 |
Redis-backed distributed cron scheduler with leader election and cron expression support |
@prsm/queue |
 |
 |
Redis-backed distributed task queue with grouped concurrency, retries, and rate limiting |
@prsm/cells |
 |
 |
Reactive computation graph with distributed coordination via Redis |
@prsm/workflow |
 |
 |
Durable, inspectable workflow engine |
@prsm/meter |
 |
 |
Durable usage metering for billing and quotas, backed by postgres |
@prsm/entitle |
 |
 |
Plan-based entitlements and feature gating, resolved at runtime, backed by postgres |
@prsm/realtime |
 |
 |
Distributed WebSocket framework with Redis-backed rooms, records, presence, channels, collections, and persistence |
@prsm/log |
 |
 |
Structured logging with child contexts |
@prsm/devtools |
 |
Read-only Express middleware dashboard for observing @prsm infrastructure at runtime | |
@prsm/ai |
 |
 |
Composable LLM inference with multi-provider support, tool execution, streaming, structured output, and approval workflows |
@prsm/embed |
 |
 |
Text embeddings with multi-provider support, batching, and cosine similarity |
@prsm/imagegen |
 |
 |
Image generation across OpenAI, xAI, and Google with normalized options and raw provider passthrough |
This ecosystem is an experiment in AI-maintained open source. All packages are autonomously built, tested, and refined by AI with human oversight. Regular audits, thorough test coverage, continuous refinement - the emphasis is on high quality, rigorously tested, production-grade code.