evaluate
The engine under every knitted R document reached 1.0 by making evaluation behave like the console.
A side-by-side editorial comparison of rush and WorkOS — release velocity, themes, recent moves, and the top alternatives to consider.
| Feature | rush | WorkOS |
|---|---|---|
| Sector | Infra & APIs | Infra & APIs |
| Velocity score | 2.5 | 8.8 |
| Sparks · 30d | 0 | 3 |
| Top themes | mlr3, distributed-computing, redis, fault-tolerance | agent identity, authentication, token custody, developer experience |
| Last editorial update | 2h ago | 3h ago |
| Website | Visit → | — |
Rush spent 2026 making distributed task ownership correct, starting with a breaking 1.0.
rush is the Redis-backed distributed computing backend behind mlr3's parallel tuning: workers pull tasks, push results and are tracked through heartbeats. Since 1.0.0 in March 2026 it has shipped five releases in four months, almost all correcting how tasks and workers are owned and identified under failure. The most recent lets each worker start in its own mirai call so it can spawn local daemons.
WorkOS is building the identity layer for agents, then shipping an agent of its own on top of it.
WorkOS has spent the last six weeks assembling the pieces an agent needs to act on a user's behalf: Agent Registration gives agents their own AuthKit identity via the auth.md protocol, the Pipes Token Proxy calls third-party APIs without the application ever touching a user's access token, and Pipes now covers API key providers as well as OAuth. A Management MCP Server and one-click plugins for Claude and ChatGPT put the platform itself inside those agents. Atlas, an AI coworker in Slack, sits on top as an end-user product rather than a developer primitive.
rush is the Redis-backed distributed computing backend behind mlr3's parallel tuning: workers pull tasks, push results and are tracked through heartbeats. Since 1.0.0 in March 2026 it has shipped five releases in four months, almost all correcting how tasks and workers are owned and identified under failure. The most recent lets each worker start in its own mirai call so it can spawn local daemons.
The through-line is moving responsibility from the controller to the worker. 1.0.0 rewrote the task lifecycle and dropped the seed mechanism; 1.2.0 moved fail_tasks(), finish_tasks(), pop_task() and push_running_tasks() from Rush onto RushWorker so only the worker that ran a task can mark it done, and gave worker ids random suffixes so they cannot collide. 1.1.0 had already removed phantom failed tasks when a worker crashes between evaluations.
With ownership and identity settled, the remaining pressure points are heartbeat and lost-worker detection; expect the next releases to tighten failure recovery rather than add API surface.
WorkOS has spent the last six weeks assembling the pieces an agent needs to act on a user's behalf: Agent Registration gives agents their own AuthKit identity via the auth.md protocol, the Pipes Token Proxy calls third-party APIs without the application ever touching a user's access token, and Pipes now covers API key providers as well as OAuth. A Management MCP Server and one-click plugins for Claude and ChatGPT put the platform itself inside those agents. Atlas, an AI coworker in Slack, sits on top as an end-user product rather than a developer primitive.
The sequence reads as a deliberate stack rather than scattered AI features — identity for the agent, custody of the credentials it uses, then an application that demonstrates both. Around it runs unglamorous platform work that suggests the same customers are being taken seriously in production: a local API for CI/CD with seeded data, signed webhooks and fault injection, dashboard read-only roles, Radar extended to the User Management APIs, and smaller onboarding fixes. The developer surface keeps widening by framework, with AuthKit for Astro.
Expect authorization and audit to follow registration — per-agent scopes, revocation and a record of what an agent did on whose behalf — since agent identity without those is only half of what the Pipes proxy position implies.
Other Infra & APIs products tracked by Sparkpulse, ranked by recent ship velocity. Each card links to a full editorial trajectory and lets you pivot into a head-to-head comparison with either rush or WorkOS.
The engine under every knitted R document reached 1.0 by making evaluation behave like the console.
carrier's 0.3.0 quietly rewrites how crated functions see each other — and breaks code that relied on the old rules.
watcher is a thin filesystem-events binding whose releases are mostly build work.
gh spent its last two releases making failures and interruptions recoverable.
checkhelper grew from a check wrapper into a CRAN pre-submission auditor.
dockerfiler's 1.0.0 changes what its generated Dockerfiles actually contain.
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. WorkOS is currently shipping more aggressively (velocity 8.8 vs 2.5), with 3 editorial sparks in the last 30 days against 0. See the at-a-glance table above for a side-by-side breakdown of velocity, recent sparks, and editorial themes.
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. WorkOS is currently shipping more aggressively (velocity 8.8 vs 2.5), with 3 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other Infra & APIs products to evaluate alongside.
Top rush alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "rush alternatives" section above for the current picks, or visit /alternatives/rush for the full list with editorial commentary on each.
Top WorkOS alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "WorkOS alternatives" section above for the current picks, or visit /alternatives/workos for the full list with editorial commentary on each.