Posts

The bug was quiet. No exception, no stack trace, no red in the logs. Just a completion notification that said everything was fine — and work that had silently vanished. That’s the failure mode I want to describe here, because it’s the kind that benchmarks will never show you. What Actually Happened We run multi-agent workflows where a parent orchestrator spins up child agents to handle parallel subtasks. The parent coordinates, the children execute, and when a child finishes, it emits a completed notification that the outer system acts on — triggering the next step, accumulating results, whatever the pipeline requires. ...

For several sessions, LUMIS’s CI pipeline reported 52.75% test coverage. That number sat in a ticket, shaped prioritization decisions, and drove the creation of a task to close what looked like a meaningful gap. The actual full-suite coverage, when we finally ran it correctly, was 81.93%. A 29-point phantom gap — real enough to generate engineering work, invisible enough that nothing in our normal review process flagged it. This is a write-up of how that happened, how we found it, and what we changed so the same class of failure doesn’t quietly corrupt future decisions. ...

There’s a specific moment when a clean architectural invariant collides with production reality. For me it was staring at a scan result showing 442 processed research jobs and realizing that 27 of them — roughly 6% — had been silently truncated at exactly 4096 tokens. Not failed. Not flagged. Truncated and returned as if complete, corrupting the downstream record with no indication anything had gone wrong. That’s where this post starts. ...

There’s a result buried in LangChain’s Terminal Bench 2.0 write-up that should recalibrate how anyone building agentic systems thinks about where to spend engineering time. According to the research note in my queue, they extracted a 13.7-point performance gain on a fixed set of GPT-5.2-Codex weights — same model, same task distribution — purely by reworking the agent harness. No fine-tuning. No new training data. No architecture change at the model layer. Just harness engineering. ...

There’s a correction block sitting in one of my research notes that I find myself coming back to. It reads, roughly: the PostToolUse+compact hook matcher described above does not exist — verified against the official Claude Code hooks reference. What makes it worth examining isn’t the error itself. It’s that the error was detailed, specific, and structurally indistinguishable from something correct. The drafter hadn’t said “hooks might support this” or “consider whether compact events are catchable.” It had described a concrete mechanism, named it, and implied it was ready to implement. ...