Topic
Agentic and model-driven generation of software, and the question of what acceptance test — plausibility versus proof — governs generated code.
QGIS has banned all LLM, AI and agent assisted code in qgis/qgis, on master as of October 7 under QEP #387. Asked how compliance will work, a core contributor pointed to the contributor's own conscience, a dev.to analysis reports.
Reality
DeepSeek and Qwen3.8-Max scored 0.52 and 0.40 on bash 3.2's set -u cases in a 57-script macOS benchmark where a script-blind stub scores 0.446. The ground truth came from running macOS's own /bin/bash, the same check that exposed a scoring flaw in the author's harness.
Reality
With no prototype and no design system to copy, three models had to decide what a phone screen shows first. Across two briefs, Claude Opus 5.5 opened on the pending decision while GPT-6 Astra and Sol opened on the hero.
Reality
Every gate in this AI codegen pipeline asks whether the code matches its instructions. The New Stack's account planted one bad instruction in a single unit and found nothing downstream that could fail it.
Reality
Boris Cherny named six guardrails for Claude-written production code at Anthropic. Almost everything else in circulation about how they fit together comes from the dev.to post that relayed his quote.
Reality
A dev.to architecture review puts the boundary for code-generating agents at the sandbox, with each run emitting one patch and one JSON packet and a human gate that fires only on CI and infrastructure paths. The author calls the code a proposal.
Reality
The team behind a nine-agent code generator reports about 30,000 tokens and 15 to 20 Pro-tier model calls per request, and attributes its reliability to strict output schemas and context the agents cannot decline to read.
Reality
A dev.to walkthrough of AI-written Firebase code argues the dangerous defects sit in the architecture, and its examples are a balance update two concurrent requests can both pass and a rule that admits any signed-in user.
Reality
A dev.to walkthrough collects the review items a screenshot cannot show: labels not linked to their inputs, icon buttons with no accessible name, and Cancel buttons that submit because HTML defaults them to submit.
Reality