Consequence from the field guide: full tracing from the first commit. The remediation loop is observe a failure → capture its trace → convert it into an eval case → change an input-boundary → verify the eval catches it. That loop — failures flowing back into the eval set — is the actual engine of
Consequence from the field guide: default to the least dynamic structure that solves the problem. Chain first. Add routing when inputs are heterogeneous. Graduate to a loop only when you can name the specific dynamism that demands it. Every loop in your architecture should have to justify its own existence.
Consequence from the field guide: constrain the executor's action space hard — few, well-scoped, ideally reversible or dry-runnable tools. Draft-commit the irreversible ones. Set hard budgets. Checkpoint. You are trading the agent's freedom for your ability to recover. How it's done A
Consequence from the field guide: measure the retriever independently of the generator; track recall/precision of retrieval on its own; and make "I don't have that information" a first-class, explicitly-rewarded behavior — because the default over a gap is a confident lie. How it'
Consequence from the field guide: assume any content the agent reads may be hostile. Design the harness so the worst an injection can do is bounded by the capabilities you granted — and grant few. This is a containment problem, not a fix. How it's done You cannot separate
Consequence from the field guide: eval-first, in CI, against a golden set, from the first commit. Prefer programmatic gates. Treat the judge as an instrument to calibrate, never as an oracle to trust. With EU AI Act enforcement arriving August 2026, this becomes an audit trail, not just hygiene.
Consequence from the field guide: treat compaction as an encoder-design problem. Decide explicitly what must survive a pass — decisions, constraints, identifiers, open questions — drop the rest, and measure the reliability delta on your evals when you change the policy. You are doing rate-distortion tradeoffs, so measure the distortion.
An engineer handed "book 30 minutes with Sarah tomorrow afternoon" produces a correct calendar entry in what feels like one step. It isn't one step. It's five distinct guarantees, established simultaneously because a single semantic model of the task produces all of them at
An engineer who has worked on your project for three months carries it without effort. They know the deploy slipped from Tuesday to Thursday, that you rejected Mongo in week two and why, that you're now in Pune rather than Bangalore, and that the connection-pool number someone
The bimodal code strategy has become the industry's consensus answer to AI-generated code risk. The argument goes: treat AI code generation as a spectrum. Automate aggressively where verification cost is low — boilerplate, scaffolding, tests. Require human authorship where verification cost is high — payment logic, security boundaries, mission-
For thirty years the job has been the same: make unreliable components — networks, disks, nodes — compose into reliable systems. The components failed in understood ways. They crashed, they lagged, they partitioned, they occasionally corrupted a byte. We built an entire discipline to tame that: timeouts, retries, quorums, replication, idempotency keys,
There is a simple test for whether a software system is doing something hard. Look at the questions it has to answer. If most of them have the shape "what is the value of X?", the system is doing lookups. If most of them have the shape "