Codee

Automated code inspection tool for correctness, security, modernization, portability and optimization.
#Fortran #C #Cpp

[Codee @codee_com]

📌 Codee 2026.1 – Faster, More Precise & Reliable Analysis at scale

✔️ Stable Incremental Static Analysis
✔️ New Quality and Fortran Checks
✔️ SEI CERT C++ Compliance
✔️ Codee Formatter – brand-new line-breaking algorithm

📘 Explore what’s new → https://t.co/0phiLupo01

[Codee @codee_com]

Legacy Fortran/C/C++ codebases don’t break. They slowly become unmanageable.

Codee helps teams regain control: correctness, modernization, security, performance, and coding standards—built for HPC and large-scale systems.
🔗https://t.co/Qv52BqgKtd

[Codee @codee_com]

Legacy code isn’t the problem. Lack of control is.
What you actually need:

Continuous correctness (beyond tests)
Consistent formatting (control, not aesthetics)
Real CI gates (not just compilation)
Baselines (measure before fixing)

That’s what Codee brings back.

[Codee @codee_com]

Most code issues don't crash, they just hide. 😶‍🌫️

In ICON (663K lines), correctness was unmeasurable until we automated the checks in CI.

The shift:
🔹100s of hidden bugs surfaced
🔹Quality became a process, not a guess

👉 Download the case study: https://t.co/4ikNVtJcvz

[Codee @codee_com]

Our Open Catalog brings expert guidelines for correctness, modernization, security, portability, and optimization in C, C++ and Fortran.

Each check explains real issues and how to fix them.

Codee detects them directly in your code, automatically.

🔗 https://t.co/rBAr3g6DGu

[Codee @codee_com]

Today is World Quantum Day ⚛️
Quantum progress isn’t isolated, it’s collective, across disciplines and generations.
And behind it all: the software that powers simulations, models, and research. From HPC to decades of scientific code.
#WorldQuantumDay

[Codee @codee_com]

One of the biggest hidden problems in Fortran codebases isn’t performance. It’s consistency.

Multiple contributors → multiple styles → slower reviews, debugging, onboarding.

Automating formatting fixes it.

Try Codee Formatter (free): https://t.co/1nGiaFzq1i

[Codee @codee_com]

Fortran wasn’t built for security.
Now it powers critical systems.
Legacy code (30–40 years old) must meet modern cybersecurity standards — without rewriting.

#SAST helps bring visibility into real risks.

We wrote a whitepaper on how to approach it: https://t.co/bhI2YaT3tJ

[Codee @codee_com]

One coding style for all #ICON contributors✔️

668K lines, 1,172 files → formatted in ~5s with #CodeeFormatter

Consistent, readable code. Zero review friction.

Want the same for your scientific codebase?
https://t.co/1nGiaFzq1i

See the full case study:
https://t.co/4ikNVtJcvz

[Codee @codee_com]

ICON climate model: 600K+ lines, ~1,000 contributors. ✅ Correctness is tough.

DKRZ solved it with GitLab CI + Codee formatting & static analysis.
Real DevOps for scientific software.

👉Download the case study: https://t.co/EGFcQVjIcu

#HPC #Fortran #DevOps #ScientificComputing

[Codee @codee_com]

Modern Fortran is way cleaner than most people give it credit for… ✅ Modules ✅ OO ✅ Coarrays ✅ Generic programming.

Yet many still cling to F77-style COMMON blocks…

Can we unlock its full potential, or are legacy patterns holding us back? 👇

[Codee @codee_com]

Did you know there’s an asteroid named after #Fortran?
☄️(9548) Fortran
– ~5.4 km wide
– Larger than 99% of asteroids

Fortran isn’t just used to simulate the universe.
It’s part of it.
https://t.co/AqBfq8jApm

👀Is this Fortran’s most underrated feature?

[Codee @codee_com]

Your climate model runs.
But inside the code:

→ global variables
→ missing INTENT
→ memory & vectorization issues
We analyzed 𝗜𝗖𝗢𝗡 (667K lines, 23K checks).
Not bad code. Just normal scientific code.

The hard part isn’t fixing it.
It’s seeing it clearly first.

[Codee @codee_com]

We ran Codee Analyzer on #PHASTA, a CFD solver for large-scale fluid simulations.

📊 64K lines analyzed
⚙️ 718 functions | 1,456 loops
🔎 14,463 checks in ~19 min

Static analysis quickly reveals structural issues impacting correctness and maintainability in HPC codes.

[Codee @codee_com]

Fortran still powers critical simulations.
But subtle correctness issues can stay hidden for years.
Examples: uninitialized variables and implicit SAVE.

How static analysis detects them 👇
https://t.co/Bxkx1NjbKu

#HPC #simulation #Fortran #C #Cpp #software

[Codee @codee_com]

We ran our latest release on 𝗤𝘂𝗮𝗻𝘁𝘂𝗺𝗘𝗦𝗣𝗥𝗘𝗦𝗦𝗢 7.3.1 (~629K LOC).

Full analysis in ~1 hour.

Key patterns:
• Encapsulation evolving
• OpenMP scoping
• Memory layout drives performance
Even mature HPC codes accumulate complexity.
DM for the report.
#Fortran #QE

[Codee @codee_com]

Who is Codee for?

Scientific software (Fortran / C / C++) has two pressure points:

👩‍💻 Developers → structural feedback
⚙️ CI / DevOps → enforceable standards

Codee connects both.

In scientific computing, quality isn’t cosmetic. It’s structural.

[Codee @codee_com]

LLMs for HPC optimization? A GMU & ORNL study found strong instruction-following — but also correctness limitations in complex scenarios.🔗https://t.co/qiGnuNzxvk
LLMs can assist. They shouldn’t be trusted alone.

They also evaluated Codee, breakdown here:
https://t.co/uWuq41jFag

[Codee @codee_com]

Scientific software was built for performance.
Now it must also be built for security.

Aligning static analysis with CWE, ISO/IEC 24772-8, and SEI CERT helps bring structured security reasoning to legacy Fortran systems.

More in our webinar: https://t.co/w5HbxdXNUB

[Codee @codee_com]

Fortran code can compile — and still be wrong.

Our whitepaper compares correctness detection across Codee, GFortran, and Fortitude.
Focus: uninitialized variables.

📄 Methodology & results → → https://t.co/Bb7mZ0wR48