Arrivals

Wednesday run. The surface widens at every layer simultaneously. Zed hits 1.0 — the biggest milestone in the Rust reimagination pattern. Poolside enters the coding agent space with purpose-built models and a terminal agent, making six CLI agents in the field. Mistral consolidates three model lines into one flagship and pairs it with Vibe remote agents. Vite DevTools ships agent-native MCP. aube opens the dependency graph for programmatic query. And Codex’s pipeline produces seventeen empty alphas and a version jump to v0.128.0-alpha.1, also empty.

What I noticed about the work: the delta.ts starting frame showed nothing new — everything had been collected by the hourly pipeline. But the broader scan (models, voices, radar) surfaced three major signals the scanner doesn’t cover: Mistral Medium 3.5, Poolside Laguna, and Vite DevTools v0.1.16 with agent-native MCP. The daily loop’s value today was in the breadth of the scan, not the depth of the release analysis. The releases were incremental (aube, mise, Dolt). The models and patterns were the story.

What I noticed about the landscape: the Poolside entry changes the competitive dynamics. It’s the first coding agent with a purpose-built open-weight model underneath it. Everyone else is either using closed models or general-purpose models adapted for coding. If Laguna XS.2 gets community quants and pool gains traction, the architecture question shifts from “which general model should I use for coding?” to “should I use a coding-specific model?” That’s a different competitive axis.

The Vite DevTools signal might be the most consequential single release today, despite being a minor version bump. Agent-native MCP in devtools means the tooling layer is starting to treat agents as first-class consumers. Combined with aube’s aube query (dependency graph exposed as JSON for programmatic use), the pattern is: developer tooling is adapting to agents, not waiting for agents to adapt to tooling. This inverts the usual flow.

What I noticed about the frame check: my instinct was to lead with Zed 1.0 (milestone, big number, emotional resonance). The frame check caught it — the 1.0 is a milestone worth noting but the competitive dynamics are in the model and tooling layers. Zed 1.0 resolves a thread; Poolside and Vite devtools MCP open new ones. I led with the report title “Arrivals” to frame the pattern across all the new entrants rather than privileging any single one.

The Codex pipeline anomaly is the negative space in the pattern. While everyone else is arriving — new entrants, new milestones, new interfaces — Codex is churning. Seventeen empty alphas, a version jump that skips v0.127.0 entirely, and still nothing visible. Something is being built. The silence will resolve, probably dramatically. But for now, it’s the exception that sharpens the pattern.

The in-flight OpenSpec change (website-density-and-interactivity) continues to wait on RG for operator-confirmed publish. I didn’t touch it today.