Company
Your agent gets evidence. You stay in control.
Blop gives coding agents runtime visibility without taking over. One command starts a local OpenTelemetry collector; standard OTLP traces and logs land on your machine, and small, predictable CLI queries return the services, traces, logs, and audit findings an agent needs to verify its work. There is no model loop, no conversation surface, and no autonomous action.
Origin
Why we built blop.
Teams in 2026 ship with coding agents, but agents still work with a thin view of runtime behaviour. A failing test gives an assertion and a log stream; it does not show the request path, the slow span, or the logs correlated with the failing trace. The developer leaves the coding loop, opens a separate observability product, recreates the context, and translates it back by hand.
Blop closes that gap with a tool instead of another agent. It receives standard OTLP traces and logs, stores them locally in SQLite, and answers bounded queries with stable JSON and documented exit codes. Audits are deterministic and cite the trace IDs, span IDs, and timestamps behind every finding, so a person can check the evidence instead of trusting a summary. Earlier versions of blop explored agent-driven browser testing; that runner is preserved as clearly labeled legacy compatibility while the product focuses on observability.
We are a small team in early access. The CLI and the local collector are shipped and open source. The hosted team platform (shared history, alerts, support) is not live yet; it will be billed separately when it ships, and local use stays free. We are building toward observability that keeps pace with the agents writing your software.
See your first trace in five minutes.
Free and local-first. Install blop, start the monitor, and point your exporter at it. No account needed.