Amazon's Verus Tool Mathematically Proves Rust Code Correct for Every Input

·8 views

Amazon has published details about Verus, a formal verification tool used in its Firecracker and AWS Lambda infrastructure that mathematically proves annotated Rust functions correct against a specification for all possible inputs. Unlike traditional code review, Verus turns correctness into a binary pass/fail result rather than a human judgment call. The article contrasts this with AI-based code review, which lacks a single objective, measurable signal needed to train models effectively via reinforcement learning. A separate experiment showed a 4-billion-parameter model improved Postgres query planning by 44.7% on join-heavy queries precisely because execution time provided a clean, verifiable reward signal. The author argues that scaling AI code review depends not on building better reviewer models, but on expanding the share of review questions that tools can formally verify rather than subjectively judge.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Developer Demonstrates Sending Raw TPM 2.0 Commands via Windows TBS in Rust

A developer has published a technical walkthrough on using Windows TPM Base Services (TBS) to send raw TPM 2.0 commands directly through Rust code. The guide covers the full TPM2_GetRandom workflow, including creating a TBS context, marshalling the command, submitting it, and parsing the response to retrieve random bytes. TPM2_GetRandom was chosen as the example because it requires no session and has minimal parameters, making it one of the simpler TPM commands to implement. The tutorial relies on the windows-sys crate and highlights key details such as big-endian byte encoding for TPM command headers. Notably, the TPM may return fewer random bytes than requested if it cannot generate the full amount in a single operation.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Flutter 'LEGO Architecture': Writing Modular, Reusable Code Like Interlocking Bricks

Developer Atuoha Anthony published a detailed guide on freeCodeCamp on September 11, 2026, explaining how to structure Flutter code using a concept called LEGO Architecture. The approach treats each widget, class, or service as a LEGO brick with a single responsibility and a standardized interface — called a 'stud' — that allows it to connect with other components without knowing their internal details. A common Flutter widget like Padding is used as a practical example, since it handles only spacing and accepts any widget as a child, making it inherently reusable. The guide contrasts this with tightly coupled widgets, such as a PriceTag that simultaneously manages both visual styling and content, which limits reuse and forces code duplication. Atuoha also emphasizes that good modular design should allow behavior to be swapped out — for example, a cart button should not need to know whether it calls a REST API, writes locally, or logs to a console.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Harness Engineering: Why AI Context Design Is Now a Core Dev Responsibility

Harness Engineering refers to the practice of building structured environments that control how AI models access tools, context, and resources — going beyond simply choosing which model to use. A real-world example illustrates the cost of ignoring it: a product manager querying ChatGPT about a production bug received a hallucinated table name because the model had no schema context, wasting significant debugging time. A well-designed harness addresses this through steering files, targeted MCP servers that return only relevant data, and model routing — one team observed token usage drop from roughly 50,000 to 15,000 on a specific task after improving context precision. Mature harness setups also include granular permissions, mandatory human-in-the-loop checkpoints for sensitive actions, and full observability with audit trails and cost metrics. The article argues that developers who understand harness design have a responsibility to propagate it across their organisations, equipping non-technical colleagues like PMs with safe, context-aware AI tooling rather than leaving them to rely on generic chat interfaces.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

AI Speeds Up Drafting But Coordination Bottlenecks Still Delay Business Workflows

AI tools can rapidly generate draft responses to complex business documents like security questionnaires, but the overall process often stalls due to unresolved approvals, unclear ownership, and coordination gaps that drafting assistance cannot fix. A faster draft can actually worsen bottlenecks by flooding a small group of reviewers with more material to check alongside existing workloads. The core issue is that completing a business process requires decisions, verified evidence, authorized sign-offs, and proper handoffs — steps AI currently struggles to replace. Experts argue that AI investments should be evaluated against the full workflow rather than just the drafting stage, since a populated document is an intermediate step, not a finished deliverable. Before adopting new tools, organizations are advised to audit recent workflows to identify exactly where delays occurred and what finally resolved them.

0 comments Read more at DEV Community