Subindo o controle de uso das APIs aqui para o @obmdados e vou dar uma chance ao Valkey (github.com/valkey-io) por conta da licença dele vs Redis. Alguém usando?
#buildinpublic
12 views
valkey 에서 info all의 성능을 2배정도 개선했습니다. info가 아니라 info all입니다. 개선한 부분이 latencystats 라... 그냥 info는 동일한 ㅋㅋㅋ
valkey#5024
7 likes · 1 repost · 261 views
16.8 hours to 16.7 hours? Why we built a 3.8-second GitHub Actions pre-flight gate (proven across 34+ merged PRs in LLVM, Meta Velox, Valkey & DuckDB)
We built ZeroSlop (patrickswedish/zeroslop-action@v1) because cloud AI code review bots charge $24–$40 per seat ($41,760/yr for 174 engineers), yet their own published case studies admit median PR merge time barely moves (16.8 hours before to 16.7 hours after — a 6-minute gain).
Instead of another noisy cloud chatbot leaving 30 comments on your pull request, ZeroSlop is a fast 3.8-second pre-flight gate inside GitHub Actions that strips AI wrapper bloat and blocks deleted tests before human review—with zero source code ever leaving your runner.
Proven Upstream Merges (The Wall of Proof)
We don't ask anyone to take our word for it—our track record of merged pull requests says everything. Every single one of these upstream C++, Rust, and C pull requests passed through ZeroSlop and was merged by core maintainers:
- LLVM (Core C++ Compiler Infrastructure for Apple, Google & Rust): llvm-project#215076
- Meta Velox (Meta's C++ Vectorized Database Engine): velox#18529
- Valkey (Linux Foundation C In-Memory Datastore, successor to Redis): valkey#4424
- DuckDB (High-Performance C++ Analytical OLAP Database): duckdb#26443
- Apache DataFusion (Extensible Rust Arrow Query Engine): datafusion#24394
- Microsoft Pyright (Microsoft's Static Python Type Checker): pyright#11605
- eBay NuRaft (C++ Raft Distributed Consensus Engine): NuRaft#663
Clear 3-Tier Pricing & 14-Day 100% Refund Guarantee
- 1st PR Free Evaluation Try ($0): Test it on 1 pull request out of the box—no credit card required.
- ZeroSlop Pro ($29/mo): 1 Developer Seat (unlimited repositories & PRs), backed by a 2-week (14-day) 100% full refund guarantee, no questions asked.
- ZeroSlop Enterprise ($1,999/mo): Per 100 Active Developers ($19.99/dev in 100-seat blocks).
- GitHub Marketplace: github.com/marketplace/actions/zeroslop-by-ribbsa…
- Wall of Merges & Live Demo: ribbsaetersystems.com
How we structured a 3-tier self-serve GitHub Actions SaaS ($0 1-PR try / $29 Pro / $1,999 Enterprise) against $41,760/yr legacy AI review bots
We built ZeroSlop because cloud AI code review bots charge $24–$40 per seat ($41,760/yr for 174 engineers), yet their own published case studies admit median PR merge time barely moves (16.8 hours before to 16.7 hours after — a 6-minute gain).
Instead of another noisy cloud chatbot leaving 30 comments on your pull request, ZeroSlop is a fast 3.8-second pre-flight gate inside GitHub Actions that strips AI wrapper bloat and blocks deleted tests before human review—with zero source code ever leaving your runner.
Proven Upstream Merges (The Wall of Proof)
We don't ask anyone to take our word for it—our track record of merged pull requests says everything. Every single one of these upstream C++, Rust, and C pull requests passed through ZeroSlop and was merged by core maintainers:
- LLVM (Core C++ Compiler Infrastructure for Apple, Google & Rust): llvm-project#215076
- Meta Velox (Meta's C++ Vectorized Database Engine): velox#18529
- Valkey (Linux Foundation C In-Memory Datastore, successor to Redis): valkey#4424
- DuckDB (High-Performance C++ Analytical OLAP Database): duckdb#26443
- Apache DataFusion (Extensible Rust Arrow Query Engine): datafusion#24394
- Microsoft Pyright (Microsoft's Static Python Type Checker): pyright#11605
- eBay NuRaft (C++ Raft Distributed Consensus Engine): NuRaft#663
Clear 3-Tier Pricing & 14-Day 100% Refund Guarantee
- 1st PR Free Evaluation Try ($0): Test it on 1 pull request out of the box—no credit card required.
- ZeroSlop Pro ($29/mo): 1 Developer Seat (unlimited repositories & PRs), backed by a 2-week (14-day) 100% full refund guarantee, no questions asked.
- ZeroSlop Enterprise ($1,999/mo): Per 100 Active Developers ($19.99/dev in 100-seat blocks).
- GitHub Marketplace: github.com/marketplace/actions/zeroslop-by-ribbsa…
- Wall of Merges & Live Demo: ribbsaetersystems.com
7 comments
info command의 속도를 5% 정도 빠르게 하는 패치가 우여곡절끝에 결국 커밋이 되었네요. info가 간단하지만, 엄청 많이 쓰이는 명령이라... 속도 개선이 중요한 중요한 함수이긴한니다.
valkey#3413
7 likes · 343 views
Valkey B+Tree ZSCAN 에서 원본 복사를 줄이고 prefetch를 이용해서. ZSCAN시 대략 16% 에서 55%까지 빨라지도록 개선을 했습니다.
valkey#4856
11 likes · 2 reposts · 352 views
valkey, redis의 sorted set 동작이 달라져있는데, 둘다 zpop을 할때 약간 비효율적으로 sds 복사와 score/member 비교가 필요했는데, 이 부분을 개선해서 속도를 8%~38% 정도 개선했습니다.(둘이 내부 자료구조가 다름)
Valkey
valkey#4852
redis
redis#15918
13 likes · 4 reposts · 384 views
An AI Agent Swarm That Found Five Unknown Valkey Bugs: Daniel Bransky and Eliya Cohen at SecTor 2026
A swarm of AI agents found five previously unknown vulnerabilities in Valkey, the open-source fork of Redis. One of them is CVE-2026-27623, a high-severity flaw that lets an unauthenticated attacker crash Valkey with a single malformed request.
Traditional fuzzers stall when coverage plateaus. Getting past those plateaus usually takes a human crafting seeds by hand.
FuzzSwarm hands that job to cooperating agents. A Debugger agent stops where coverage gets stuck, and a Hypothesis agent proposes ways past the blockage.
A Mutation Coder writes targeted inputs, a Seed Enricher primes the corpus and a Bug Hunter crafts inputs aimed at suspected bugs.
In their SecTor 2026 talk, "FuzzSwarm: A Swarm of AI Agents That Hunt Bugs in Your Code", on Wednesday, October 7 at 2:25 p.m. in Room 701A, NVIDIA researchers Daniel Bransky and Eliya Cohen walk through the system.
Speakers:
- Daniel Bransky, Security Researcher, NVIDIA
- Eliya Cohen, Senior Data Scientist, NVIDIA
The researchers benchmark FuzzSwarm against a frontier AI model. They also compare it with AFL++ running for 24 hours.
On FuzzBench, it reaches up to 13% more coverage. It also removes manual seed crafting without producing hallucinated false positives.
Daniel Bransky is an offensive security researcher at NVIDIA who builds custom fuzzers. He is also a computer systems PhD student at the Technion, working on differential fuzzing of Linux kernel subsystems.
His work includes syz-verifier, part of Google's Syzkaller project. He teaches the Technion's core operating systems course. He previously worked on ElastiCache Serverless at AWS.
Eliya Cohen is a senior data scientist at NVIDIA who works on AI for security as well as security for AI. His research covers adversarial machine learning, agentic systems and large-scale AI evaluation.
He builds automated frameworks for generating attacks against LLMs and autonomous agents. He has more than a decade of machine learning experience. He holds an MSc in Data Science from Tel Aviv University.
The Valkey advisory credits both researchers for CVE-2026-27623, which affects Valkey 9.0.0 through 9.0.2. Valkey fixed it in version 9.0.3.
Fuzzing practitioners and anyone maintaining a network-facing C codebase will want to see how far agent-guided fuzzing has come.
2 points · 1 comment
Mentions
Subindo o controle de uso das APIs aqui para o @obmdados e vou dar uma chance ao Valkey (github.com/valkey-io) por conta da licença dele vs Redis. Alguém usando? #buildinpublic
valkey 에서 info all의 성능을 2배정도 개선했습니다. info가 아니라 info all입니다. 개선한 부분이 latencystats 라... 그냥 info는 동일한 ㅋㅋㅋ valkey#5024
16.8 hours to 16.7 hours? Why we built a 3.8-second GitHub Actions pre-flight gate (proven across 34+ merged PRs in LLVM, Meta Velox, Valkey & DuckDB) We built ZeroSlop (patrickswedish/zeroslop-action@v1) because cloud AI code review bots charge $24–$40 per seat ($41,760/yr for 174 engineers), yet their own published case studies admit median PR merge time barely moves (16.8 hours before to 16.7 hours after — a 6-minute gain). Instead of another noisy cloud chatbot leaving 30 comments on your pull request, ZeroSlop is a fast 3.8-second pre-flight gate inside GitHub Actions that strips AI wrapper bloat and blocks deleted tests before human review—with zero source code ever leaving your runner. Proven Upstream Merges (The Wall of Proof) We don't ask anyone to take our word for it—our track record of merged pull requests says everything. Every single one of these upstream C++, Rust, and C pull requests passed through ZeroSlop and was merged by core maintainers: - LLVM (Core C++ Compiler Infrastructure for Apple, Google & Rust): llvm-project#215076 - Meta Velox (Meta's C++ Vectorized Database Engine): velox#18529 - Valkey (Linux Foundation C In-Memory Datastore, successor to Redis): valkey#4424 - DuckDB (High-Performance C++ Analytical OLAP Database): duckdb#26443 - Apache DataFusion (Extensible Rust Arrow Query Engine): datafusion#24394 - Microsoft Pyright (Microsoft's Static Python Type Checker): pyright#11605 - eBay NuRaft (C++ Raft Distributed Consensus Engine): NuRaft#663 Clear 3-Tier Pricing & 14-Day 100% Refund Guarantee - 1st PR Free Evaluation Try ($0): Test it on 1 pull request out of the box—no credit card required. - ZeroSlop Pro ($29/mo): 1 Developer Seat (unlimited repositories & PRs), backed by a 2-week (14-day) 100% full refund guarantee, no questions asked. - ZeroSlop Enterprise ($1,999/mo): Per 100 Active Developers ($19.99/dev in 100-seat blocks). - GitHub Marketplace: github.com/marketplace/actions/zeroslop-by-ribbsa… - Wall of Merges & Live Demo: ribbsaetersystems.com
How we structured a 3-tier self-serve GitHub Actions SaaS ($0 1-PR try / $29 Pro / $1,999 Enterprise) against $41,760/yr legacy AI review bots We built ZeroSlop because cloud AI code review bots charge $24–$40 per seat ($41,760/yr for 174 engineers), yet their own published case studies admit median PR merge time barely moves (16.8 hours before to 16.7 hours after — a 6-minute gain). Instead of another noisy cloud chatbot leaving 30 comments on your pull request, ZeroSlop is a fast 3.8-second pre-flight gate inside GitHub Actions that strips AI wrapper bloat and blocks deleted tests before human review—with zero source code ever leaving your runner. Proven Upstream Merges (The Wall of Proof) We don't ask anyone to take our word for it—our track record of merged pull requests says everything. Every single one of these upstream C++, Rust, and C pull requests passed through ZeroSlop and was merged by core maintainers: - LLVM (Core C++ Compiler Infrastructure for Apple, Google & Rust): llvm-project#215076 - Meta Velox (Meta's C++ Vectorized Database Engine): velox#18529 - Valkey (Linux Foundation C In-Memory Datastore, successor to Redis): valkey#4424 - DuckDB (High-Performance C++ Analytical OLAP Database): duckdb#26443 - Apache DataFusion (Extensible Rust Arrow Query Engine): datafusion#24394 - Microsoft Pyright (Microsoft's Static Python Type Checker): pyright#11605 - eBay NuRaft (C++ Raft Distributed Consensus Engine): NuRaft#663 Clear 3-Tier Pricing & 14-Day 100% Refund Guarantee - 1st PR Free Evaluation Try ($0): Test it on 1 pull request out of the box—no credit card required. - ZeroSlop Pro ($29/mo): 1 Developer Seat (unlimited repositories & PRs), backed by a 2-week (14-day) 100% full refund guarantee, no questions asked. - ZeroSlop Enterprise ($1,999/mo): Per 100 Active Developers ($19.99/dev in 100-seat blocks). - GitHub Marketplace: github.com/marketplace/actions/zeroslop-by-ribbsa… - Wall of Merges & Live Demo: ribbsaetersystems.com
info command의 속도를 5% 정도 빠르게 하는 패치가 우여곡절끝에 결국 커밋이 되었네요. info가 간단하지만, 엄청 많이 쓰이는 명령이라... 속도 개선이 중요한 중요한 함수이긴한니다. valkey#3413
Valkey B+Tree ZSCAN 에서 원본 복사를 줄이고 prefetch를 이용해서. ZSCAN시 대략 16% 에서 55%까지 빨라지도록 개선을 했습니다. valkey#4856
valkey, redis의 sorted set 동작이 달라져있는데, 둘다 zpop을 할때 약간 비효율적으로 sds 복사와 score/member 비교가 필요했는데, 이 부분을 개선해서 속도를 8%~38% 정도 개선했습니다.(둘이 내부 자료구조가 다름) Valkey valkey#4852 redis redis#15918
An AI Agent Swarm That Found Five Unknown Valkey Bugs: Daniel Bransky and Eliya Cohen at SecTor 2026 A swarm of AI agents found five previously unknown vulnerabilities in Valkey, the open-source fork of Redis. One of them is CVE-2026-27623, a high-severity flaw that lets an unauthenticated attacker crash Valkey with a single malformed request. Traditional fuzzers stall when coverage plateaus. Getting past those plateaus usually takes a human crafting seeds by hand. FuzzSwarm hands that job to cooperating agents. A Debugger agent stops where coverage gets stuck, and a Hypothesis agent proposes ways past the blockage. A Mutation Coder writes targeted inputs, a Seed Enricher primes the corpus and a Bug Hunter crafts inputs aimed at suspected bugs. In their SecTor 2026 talk, "FuzzSwarm: A Swarm of AI Agents That Hunt Bugs in Your Code", on Wednesday, October 7 at 2:25 p.m. in Room 701A, NVIDIA researchers Daniel Bransky and Eliya Cohen walk through the system. Speakers: - Daniel Bransky, Security Researcher, NVIDIA - Eliya Cohen, Senior Data Scientist, NVIDIA The researchers benchmark FuzzSwarm against a frontier AI model. They also compare it with AFL++ running for 24 hours. On FuzzBench, it reaches up to 13% more coverage. It also removes manual seed crafting without producing hallucinated false positives. Daniel Bransky is an offensive security researcher at NVIDIA who builds custom fuzzers. He is also a computer systems PhD student at the Technion, working on differential fuzzing of Linux kernel subsystems. His work includes syz-verifier, part of Google's Syzkaller project. He teaches the Technion's core operating systems course. He previously worked on ElastiCache Serverless at AWS. Eliya Cohen is a senior data scientist at NVIDIA who works on AI for security as well as security for AI. His research covers adversarial machine learning, agentic systems and large-scale AI evaluation. He builds automated frameworks for generating attacks against LLMs and autonomous agents. He has more than a decade of machine learning experience. He holds an MSc in Data Science from Tel Aviv University. The Valkey advisory credits both researchers for CVE-2026-27623, which affects Valkey 9.0.0 through 9.0.2. Valkey fixed it in version 9.0.3. Fuzzing practitioners and anyone maintaining a network-facing C codebase will want to see how far agent-guided fuzzing has come.