manos-saratsis

3 ms·

1.
by manos-saratsis 2mo ago
Good breakdown. One thing this doesn't solve fully: narrow PRs make each diff reviewable in isolation, but five narrow PRs that each look fine on their own can still interact badly once they're all merged. There need to be checks
2.
by manos-saratsis 2mo ago
Fully agree on the interpretation. Changes are massive, we checked 100k+ OSS PRs and we see plenty with 1k+ lines of code. And the bigger the PR is, the longer the time is to review and the highest the chance of rejection https:/&#x2F
3.
by manos-saratsis 2mo ago
Tool from tool differs, we do a check of intent vs actual changes https://dromeas.ai/code-review
4.
by manos-saratsis 2mo ago
Very true. A few things that worked for us: 1. Split review priority by blast radius, not diff size. A 400-line PR touching one isolated module is often faster to review than a 40-line PR touching shared components or a migration. - Push th
5.
The Code Review Paradox(dromeas.ai)
1 points by manos-saratsis 2mo ago|1 comments
6.
by manos-saratsis 2mo ago
We analysed more than 600 open source repositories and amongst others we see 63x more lines per PR than per direct commit. When teams used to review a handful of changes to stay up to day, now they understand, test and review 1000s lines of
7.
by manos-saratsis 2mo ago
Spot on and nice complement to the pre-merge half of this problem. Silent logic bugs slips through when a diff "looks right" — what you're building catches them once they're live.
8.
by manos-saratsis 2mo ago
An AI-assisted team. There are many ways to prevent issues from the dev laptop, to team code review, to pre-build test, to automated agentic testing, monitoring, canary deployments with auto-remediation but if all fail teams are the ones to
9.
by manos-saratsis 2mo ago
Very true Sylvain. The evolution of the code review agents is on have configurable autonomy, work on the human learning side, prevent issues both on pull requests or trunk based development depending your style. Working on it https:/&
10.
by manos-saratsis 2mo ago
Manually yes, PRs are often huge lately making it impossible to go through them. Code review bots help going through them. Some interesting stats here https://dromeas.ai/blog/pr-vs-trunk-what-code-review-actuall...
11.
Code review powered by an LLM council(dromeas.ai)
2 points by manos-saratsis 3mo ago|1 comments
12.
by manos-saratsis 3mo ago
With today challenges with code reviews here our point of view: multi-model code review with built-in cross-validation for security and compliance findings avoiding single-model blindspots. Different modes from assisted review to autonomous
13.
Fresh Open Claw Documentation
1 points by manos-saratsis 7mo ago|0 comments
14.
Communication Coach in Slack(mculture.xyz)
1 points by manos-saratsis 2y ago|1 comments
15.
by manos-saratsis 2y ago
I'm Manos, co-founder of Micro Culture. As a remote-hybrid digital worker I personally had challenges with chat communication. We often chat when we are stressed, when distracted during meetings and sometimes when we are on transit. Ev