Hi HN! I’m the maker of Revieko — a GitHub App that highlights architecture drift hotspots in pull requests and posts the results where reviews happen.
What you get in each PR:
- a PR comment with the top hotspots and file/line pointers
- a status check
— a full report - markdown/ json
How it works (high level): after install it builds a baseline from the repo, then compares each PR to highlight structural changes that are unusual for that codebase
What it isn’t: not a linter, not a security scanner — the goal is “signal for review”, not more noise.
It’s free right now while we learn. I’d love feedback on 2 things:
1. what would make a hotspot truly actionable in your repo?
2. what would be noise that you’d want filtered out?
One big pain point in large codebases is that LLM agents make excellent local code or architecture decisions, but they might not fit the global architectural context and rules. So local optimality often leads to subtle architecture drift. Maybe Revieko could become a critical layer in agent workflows and act as a “structural conscience” for code written by agents.
Hi HN! I’m the maker of Revieko — a GitHub App that highlights architecture drift hotspots in pull requests and posts the results where reviews happen. What you get in each PR: - a PR comment with the top hotspots and file/line pointers - a status check — a full report - markdown/ json How it works (high level): after install it builds a baseline from the repo, then compares each PR to highlight structural changes that are unusual for that codebase What it isn’t: not a linter, not a security scanner — the goal is “signal for review”, not more noise.
It’s free right now while we learn. I’d love feedback on 2 things: 1. what would make a hotspot truly actionable in your repo? 2. what would be noise that you’d want filtered out?
Install: https://github.com/apps/revieko-architecture-drift-radar Sample output: https://synqra.tech/revieko#modal-sample
One big pain point in large codebases is that LLM agents make excellent local code or architecture decisions, but they might not fit the global architectural context and rules. So local optimality often leads to subtle architecture drift. Maybe Revieko could become a critical layer in agent workflows and act as a “structural conscience” for code written by agents.