Blog

Engineering notes about AI agent observability

How to connect model activity to local processes, files, network calls, and resource use.

Performance methods · August 2026

How much overhead does AgentSight add?

Read the published 2.9% average runtime-overhead result with the raw run variability, benchmark limits, and a method for measuring your own workload.

Engineering note

Application logs explain declared agent steps; system activity shows the commands, files, processes, and network effects that actually ran.

Workflow note

Turn a large agent event stream into a bounded artifact that explains decisions, validation, side effects, and open questions.

Systems deep dive · August 2026

A source-level guide to OpenSSL, BoringSSL, rustls, stripped binaries, uprobes, and the plaintext hook points behind modern AI agent CLI traffic.