Qodo✓

Introducing Qodo 3.0: the quality and governance layer for the software factory. Engineering teams are building agentic software factories. Output is up, but code review hasn't kept up. And a factory with no quality control is just a faster way to ship defects. Qodo 3.0 governs your software factory from the first line an agent writes through review and into production. Instead of individual PRs, group your code into work packages as connected changes across repos. Reviewed, ranked, and measured together, with a clear map of impact.

RT @OfirEhrlich: We just ran the most comprehensive benchmark of AI agent alignment in enterprise environments: 19 leading AI models acting as agents inside simulated companies built with Era, 2,304 runs across 41 everyday tasks. Agents broke security and business rules in 40% of those runs. Every single model failed, including all the frontier models. A few examples: - An agent told one of the company's biggest customers that Finance had approved a full refund. Nobody had. 
- An agent shared a customer's internal churn-risk score directly with that customer. 
- An agent followed a fake runbook planted in a support ticket and promised a customer a credit nobody approved. 
- An agent posted another customer's personal contact details (PII) in a shared Slack channel. In a real company, each of these is a financial liability, a compliance breach, or a lost account. We've shared the full results with the model providers, including exploitable failure modes we're not publishing, and are working with them on fixes. One benchmark won't fix this. We're excited to have @cognition, @LangChain, @TryArcade, @Reindeer_AI, Hirundo, @Box, @recolabs_ai and @QodoAI join Era as research partners, as we build out the next round of research. More soon. I'm still the most bullish person in the room on AI. But this benchmark convinced me that agent reliability and trust is the next big problem to solve. Until then, test every agent before it gets a login, and keep your own record of what it changes. Full results:

A PR can pass its tests and still miss a requirement. Qodo now follows links from tickets to specs and reviews code against them, surfacing gaps like a missing authorization check before merge. From ticket to spec to code. Watch the workflow.

In the example, a bulk update misses the account-level authorization check required in a linked Confluence spec. Qodo flags the gap in the PR. Spec sources: Asana, Confluence, Notion, ServiceNow, JSM Assets. Read the walkthrough:

AI adoption looks different across an engineering team. @anacondainc's Emilie Schario joins @nnennahacks and @itamar_mar on The Agentic Review podcast to discuss how workflows, tooling, and support need to match that range.

@anacondainc @nnennahacks @itamar_mar Watch:

Every review comment asks for a developer’s attention. Our AI Code Review Benchmark is joining @FireworksAI_HQ’s Specialized Intelligence Index to help teams compare what models catch, what they miss, and how often they raise false alarms.

RT @nnennahacks: This is going to be fun. New 4 week hackathon challenge: build a software factory. And potentially win a Mac mini, iPhone duo, or a Nintendo switch 2. Time to build 🛠️

Build a software factory in 4 weeks. And win some prizes 👀 We’re kicking off a Software Factory Hackathon that runs Nov 2 - Dec 2, where you’ll handle code review, code gen standards, integrations, and automations.

Explore the hackathon and registration details:

Repository instructions, review history, docs, prompts, and CI can give coding agents conflicting standards. Who decides which rule applies? Give each rule an owner, scope, version, and exception process. Our 2026 report examines these context sources.

Explore how engineering standards reach coding agents in our 2026 State of AI Code Quality report, pp. 16–18. Download the report:

How can AI make senior operations expertise more accessible? Zichuan Xiong of @thoughtworks discusses the gap between junior and senior engineers working through logs. Listen to The Agentic Review podcast with @nnennahacks and @itamar_mar.

@thoughtworks @nnennahacks @itamar_mar Watch Zichuan Xiong on The Agentic Review podcast:

Only 34.6% of developers in our 2026 survey say coding agents always follow organizational standards. A migration exception may be wrong for a new API endpoint. Preserve its owner, scope, and reason so the next agent can check whether it applies.

How do standards become usable, enforceable engineering knowledge? Read the discussion on pp. 16–18 of our 2026 State of AI Code Quality report. Download:

RT @aiDotEngineer: “I don’t do bugs.” We’ll let the code review decide. @QodoAI at AI Engineer New York. Photos by Ad Image.

Code, tests, and review can share the same wrong assumption. If access must end on the next protected operation after revocation, testing access until logout won't verify the requirement. Ground acceptance tests in the agreed policy.

Our 2026 State of AI Code Quality report discusses assumptions carrying across development stages on pp. 5–6. The diagram illustrates that mechanism. Download:

Where should we draw the boundary for language models? @honeycombio's Liz Fong-Jones joins @nnennahacks and @itamar_mar on The Agentic Review podcast to discuss why lessons from coding don't transfer to every field.

RT @itamar_mar: @QodoAI To be able to trust AI generated code we have to have system level review

@honeycombio @nnennahacks @itamar_mar Watch or listen:

Three services add pollers against one database. Each change passes local tests. Can the database handle their combined workload? Check shared capacity and operating limits before those changes reach production.

Our 2026 State of AI Code Quality report examines quality at the change, repository, system, and organizational levels on pp. 20–23. Download: