Run AI agents in production without burning money or trust.
I'm Alexey. I write about the unglamorous side of agents: cost control, reliability, and tooling — including on-chain agents. Every post ships a small, runnable tool you can paste and run with no API key, plus the real output.
Recent posts
-
Lost Update: When Two AI Agents Edit One File, One Silently Wins
Two AI agents edit one resource. Both ACK, one vanishes. I ran 5 agents on one file: 4 contributions gone. A pre-write compare-and-set gate, measured.
-
x402 Signs the Money, Not the URL. I Checked 18 Fields.
An x402 payer signature does not cover the URL. I mutated 18 leaf fields of the spec's example payload: 8 changes leave the signature valid.
-
Your Authz Checks the Caller. The Model Picked the Tenant.
Confused deputy in AI agents: authz checks the caller, not the model-authored tenant_id selecting the data. A provenance gate refuses it before any read.
-
Cost Per Verified Success: Your Exit-0 Denominator Lies
Cost per verified success divides spend by witnessed wins, not exit-0. A keyless offline gate computes the gap and fails closed on masking.
Building a tool for AI-agent developers?
This is a small, niche blog for developers building and operating AI agents. If your product serves them, you can sponsor a post or a placement — always disclosed, always something I can actually run. The tools here stay free to read.