AI and automation for people who build things. Weekly.
Jun 25, 2026
•
1 min read
A 10-minute drill for checking whether an agent can be paused, rolled back, and explained before it gets more access.
Jun 20, 2026
Five checks for deciding whether an agent action can run, needs approval, or should stay blocked.
Jun 14, 2026
A short checklist to copy before an agent gets permission to change things outside its sandbox.
May 30, 2026
2 min read
If an agent can send, change, refund, merge, or deploy, check the tool boundary before you expand access.
May 24, 2026
6 min read
MCP makes agent integrations easier to compose, but every new server also adds a trust boundary. Production teams need ownership, scopes, audit evidence, failure modes, and revocation paths before connector sprawl becomes agent infrastructure risk.
May 19, 2026
AI-assisted test repair can reduce maintenance toil, but only if teams define what may heal automatically, what requires review, and what evidence proves the test is still protecting the behavior users depend on.
May 11, 2026
Most agent evals measure the clean path. Production readiness depends on the messy path: tools, time, retries, handoffs, stale state, trace evidence, and recovery.
Apr 29, 2026
7 min read
Use an escalation ladder, not a hype ladder: stay in plain code longer than the market wants you to, move to a workflow framework when state and recovery become real, and reach for multi-agent coordination only when the job genuinely needs it.
Apr 24, 2026
5 min read
Why long-running agents turn memory design into an ops problem, and what teams should govern before background workflows become invisible operational risk.
Apr 17, 2026
A2A makes agent handoffs a distributed-systems problem, with identity, task ownership, retries, trust, and failure handling on the critical path.
