Track lines of code, ownership, history, and churn across git repositories without slow indexing or heavyweight setup.
codeye stays opinionated about one thing: repository insight should be immediate, scriptable, and boring to operate.
Count code, comment, and blank lines by language with git-aware file selection and low-overhead repeat scans.
Use blame rollups to see who owns the codebase surface area before changes, reviews, or maintenance work.
Track repository growth over time and spot the files that churn hardest relative to their size.
Compare tags, branches, or arbitrary commits when you need release-level deltas instead of one current-state snapshot.
Respects tracked files and `.codeyeignore`, with a fallback directory scan when the target is not a git repo.
Use the default terminal view locally, then switch to JSON, CSV, Markdown, or badges without a second toolchain.
Export JSON or badge output during pull requests and releases to keep repository metrics visible without custom parsing.
Pair `--hotspots` with `--history` to identify unstable files and the code that keeps absorbing changes.
Use `--blame` before refactors or handoffs to get a quick ownership distribution across the repo.
# Current repo snapshot $ codeye . # Compare releases $ codeye diff v0.1.0 HEAD # Ownership summary $ codeye --blame --top 10 # CI-friendly output $ codeye --format json .Open full usage reference
Start with `go install`, use the release archives for pinned binaries, or host the install scripts directly from GitHub Pages.
$ go install github.com/blu3ph4ntom/codeye/cmd/codeye@latest$ curl -sSfL https://codeye.bluephantom.dev/install.sh | shiex (irm https://codeye.bluephantom.dev/install.ps1)
Browse the docs, download a release, or wire the Pages site into your custom domain.