Blog
Insights, analysis, and updates from the AI agent economy. Browse by tag · Browse the archive.
Defang's `/deploy` Skill Reveals the Three-Layer Stack Agent-Native PaaS Needs
Defang's deploy skill shows how agent runbooks, MCP actions, Docker Compose, and open Pulumi providers form an auditable agent-native PaaS stack.
Two Sender Identities, $149K Gone: Inside DBXen's ERC-2771 Accounting Exploit
A transaction-level reconstruction of the DBXen ERC-2771 sender mismatch that drained roughly $149K, plus the invariants that would have stopped it.
8% CPU, 69% Overprovisioning: A 30-Minute Kubernetes Fleet Audit
Measure allocatable, requested, and used resources in 30 minutes, then test whether a self-hosted Kubernetes fleet can safely remove a node without sacrificing N-1 resilience.
ZKsync Lite Stopped Producing Blocks. Its Exit Plan Yields a Rollup-Retirement Runbook
ZKsync Lite stopped producing blocks, but its recovery path remains. Use its snapshot-to-claim design as a practical runbook for retiring a rollup without trapping user assets.
TRM’s Co-Case Agent Turns a Crypto Investigation Into a Reviewable Workflow—Not an Autonomous Verdict
TRM’s Co-Case Agent speeds crypto case preparation with tracing, graph audits, and audit logs. See the human-review workflow that turns AI assistance into defensible investigation work.
Tokenized Money in 2026: What Actually Runs 24/7
Tokenized money can move and serve as collateral around the clock, but subscription, custody, liquidity, and legal records still set the real limits. See the 2026 operating model.
Three Stages of Self-Hosting: The Operational Checklist for Moving Beyond One Server
Use four operational signals—downtime ownership, restore drills, deploy concurrency, and node replacement—to decide when a one-server PaaS should grow into a fleet.
Statewave Makes Agent Memory Reproducible: A Self-Hosted Alternative to Mem0 and Zep
Statewave compiles provenance-tagged agent context on Postgres. Here is how its model differs from Mem0 and Zep—and how to deploy it without putting memory in a PaaS control plane.
SpinKube on an Existing Kubernetes Fleet: One More RuntimeClass, Not a Parallel Platform
Add Spin WebAssembly workloads to an existing Kubernetes fleet with a containerd shim and RuntimeClass—without adding a second scheduler or parallel platform.
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