Buildkite
Buildkite is forcing a major agent upgrade and handing agents a shortcut through failures.
A side-by-side editorial comparison of Windsurf and Expo — release velocity, themes, recent moves, and the top alternatives to consider.
| Feature | Windsurf | Expo |
|---|---|---|
| Sector | Infra & APIs | Infra & APIs |
| Velocity score | 0.0 | 6.3 |
| Sparks · 30d | 0 | 1 |
| Top themes | agentic-coding, devin, code-review, ide | react-native, eas, mcp, ci-cd |
| Last editorial update | 2h ago | 2h ago |
| Website | Visit → | — |
Windsurf is being rebuilt as Devin's front end — the IDE now ships the agent, the CLI, and the reviewer.
Every substantive item in this window is a Devin capability arriving inside Windsurf under the existing subscription: the Devin local agent, Devin for Terminal as a Rust CLI, and now Devin Review and Quick Review for checking AI-written code. The editor's own releases are increasingly about managing agents rather than editing text, with an Agent Command Center inbox, session sorting, and switching performance. Frontier models — Opus 4.7, GPT-5.5, SWE-1.6 — are added as routine availability notes.
Expo shut down its own coding agent and kept the MCP server — it wants to be the substrate.
Two SDK releases landed in quick succession, 56 in May and 57 at the end of June, alongside steady EAS work: Maestro test insights with flake detection, automated iOS device registration in Workflows, and a general availability date for EAS Observe. On the AI side the direction reversed — the Expo MCP Server became available on the free plan in May, and in July the Expo Agent closed beta ended with the project wound down. Entry bodies in this feed are thin, so specifics beyond the headlines are limited.
Every substantive item in this window is a Devin capability arriving inside Windsurf under the existing subscription: the Devin local agent, Devin for Terminal as a Rust CLI, and now Devin Review and Quick Review for checking AI-written code. The editor's own releases are increasingly about managing agents rather than editing text, with an Agent Command Center inbox, session sorting, and switching performance. Frontier models — Opus 4.7, GPT-5.5, SWE-1.6 — are added as routine availability notes.
The IDE is becoming one surface among several onto a shared agent harness that also runs in the terminal and in the cloud, with sessions handed between them. The review tooling signals where the bottleneck has moved: once agents write most of the code, the scarce capacity is checking it, so Cognition is shipping the reviewer alongside the writer. Bundling all of it into one subscription rather than pricing per agent is the commercial posture.
Expect the agent inbox to keep absorbing IDE surface — more parallel sessions, review as a gate on agent output — and further parity between the terminal, cloud, and editor entry points.
Two SDK releases landed in quick succession, 56 in May and 57 at the end of June, alongside steady EAS work: Maestro test insights with flake detection, automated iOS device registration in Workflows, and a general availability date for EAS Observe. On the AI side the direction reversed — the Expo MCP Server became available on the free plan in May, and in July the Expo Agent closed beta ended with the project wound down. Entry bodies in this feed are thin, so specifics beyond the headlines are limited.
Expo is consolidating around the build, test, and observe pipeline it already owns and stepping back from competing on the agent itself. Making the MCP server free while retiring Expo Agent puts the company underneath whichever assistant a developer already uses, rather than asking them to adopt another one. EAS is where the commercial weight sits, and the recent releases all reduce the manual work between a commit and a working build on a device.
Expect EAS Observe's August general availability to anchor the next stretch, with AI effort staying on the MCP surface that connects outside assistants to Expo projects.
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 Windsurf or Expo.
Buildkite is forcing a major agent upgrade and handing agents a shortcut through failures.
Statsig is packaging its own workflows as skills for someone else's agent to run.
Rootly's AI stops summarizing incidents and starts investigating them.
Notification infrastructure grows a memory: workflows can now wait on the real world.
Auth infrastructure that agents can drive and developers can run locally.
Clerk is building auth for machines: API keys, a CLI agents can drive, per-verification pricing.
See all Windsurf alternatives → · See all Expo alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. Expo is currently shipping more aggressively (velocity 6.3 vs 0.0), with 1 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. Expo is currently shipping more aggressively (velocity 6.3 vs 0.0), with 1 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 Windsurf alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "Windsurf alternatives" section above for the current picks, or visit /alternatives/windsurf for the full list with editorial commentary on each.
Top Expo alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "Expo alternatives" section above for the current picks, or visit /alternatives/expo for the full list with editorial commentary on each.