Blog
Insights, analysis, and updates from the AI agent economy. Browse by tag · Browse the archive.
CNCF's 19.9 Million Developers, +28% in Six Months: The Math That Says It's Not New Growth
CNCF says the cloud native developer population jumped 28% in six months to 19.9 million, with 7.3 million now AI developers. The arithmetic says that's not a wave of new developers — it's existing developers crossing a line, and it changes how fast a platform's agent golden path has to ship.
CAPMOX Is Maturing Fast: What a Real Cluster API Provider for Proxmox Means Beyond Hetzner-Only Node Pools
IONOS's CAPMOX shipped six releases in fifteen months of active development — here's what the changelog actually shows, how it provisions a node, who else is building the same thing, and what a second CAPI provider really costs a Hetzner-only fleet.
Canine's MCP Server Can Deploy — Coolify's Still Can't: A Tool-by-Tool Audit
Canine's built-in MCP server ships one write tool — DeployProject — where Coolify's stays fully read-only. Here's the exact tool list across Canine, Coolify, and Dokploy, and what Canine's Kubernetes-native architecture actually buys over each.
Blacksmith and Depot Colocate Docker Layer Cache on NVMe: What 40x Faster CI Builds Mean for Self-Hosted Build Pipelines
Blacksmith and Depot claim 20x-40x faster Docker builds by moving the layer cache onto NVMe disk colocated with the build machine. Here's the actual mechanism, and what it means for a self-hosted PaaS that already owns its build hardware.
Windsurf Split Into Three Companies in a Week — Then the Brand Died Too
OpenAI's $3B Windsurf deal collapsed, Google took the founders for $2.4B, and Cognition bought the rest for $250M — all in one week. Eleven months later even the brand was gone. Here's why building your deploy workflow on an open protocol beats betting on any single agent vendor's survival.
The $2 Million Threshold: What Building Your Own Platform Actually Costs
A cloud bill supposedly needs to hit $2M a year before building your own platform beats renting one — except that $2M is platform-team payroll, not an infrastructure bill, and running a self-hosted PaaS on owned hardware gets the same capability for roughly a third to two-thirds of the cost.
SPIRE Needs Your Fleet's Shape in Advance. An Orchestrator Spawning Sub-Agents Doesn't Have One.
SPIRE requires every workload to be pre-registered before it can attest — a model that breaks the moment an orchestrator spawns sub-agents on demand. Here's the actual registration race, the ClusterSPIFFEID fix, and what it still doesn't solve.
What a Zanzibar-Style Relationship Graph Buys a Deploy MCP Server Over a Scoped API Key
AuthZed is pitching SpiceDB's relationship-graph authorization for AI agents. Here's the actual SpiceDB schema for a deploy/rollback MCP server, what it can check that a scoped API key can't, and whether running it is worth the operational cost.
SNCF Rebuilt Its Cloud-Native Platform on Cluster API — Here's What Changed
CNCF gave SNCF its 2026 Top End User Award for a Cluster API rebuild that cut cluster provisioning from a month to 30 minutes and production incidents by 90%. Here's the stack, the before/after numbers, and what it proves about running Cluster API at production stakes.
Subscribe
New posts land in your reader as soon as they publish. Pick a format — all three carry the same posts.
Current feeds keep roughly two days of posts so daily polling does not miss a burst. Older entries stay reachable from the feed's next-page link in readers that follow it, or from the blog archive.
Following one topic instead? Browse tags