Replies: 2 comments
-
|
First contribution here!😁 |
Beta Was this translation helpful? Give feedback.
-
|
tldr: can we think of this from a filtering, grouping, aggregating and general git history query feature? Hi @fuleinist and @iconicvenom - this is a good initiative overall I feel. Just to provoke some thought around this - what are some usecases you think will be useful from a day to day team functioning perspective?
I recommend thinking on these aspects a bit -- because as to the implementation part -- maybe it'd be a good idea to have "filters" first on the whole of commit log, and then build stats as a module on top of it. One approach I can think of is a duckdb based approach. That is - build a table of the commit log using duckdb: https://duckdb.org/docs/current/guides/snippets/analyze_git_repository Then allow any query on top of it. And maybe have some query aliases for quick access (defaults + configurable). This will give a generic reporting engine upon which a whole host of usecases can be supported from a team lead + engineer perspective. Hopefully that sheds some light on how I think of this whole problem. As to next action - I recommend refining your proposal on the problem + solution in a bit more details - so we can settle on an execution scope. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Problem
Every
git commitwith git-lrc embeds review metadata in the commit message:LiveReview Pre-Commit Check: ran (iter:N, coverage:X%). However there is no way to query or visualize this history from the CLI — developers must manually parsegit logto understand their review coverage over time.Solution
Add a
git lrc stats(and aliaslrc stats) command that:LiveReview Pre-Commit Check:lines fromgit logfor the current repogit lrc stats --last 30— last N days breakdown (default: all time)git lrc stats --json— outputs machine-readable JSON for CI dashboards or scriptingExample output:
Example JSON output (
--json):{ "total": 64, "reviewed": { "count": 47, "avg_iter": 2.3, "avg_coverage": 74 }, "vouched": { "count": 12 }, "skipped": { "count": 5 } }Use case
Developers and teams want to understand review coverage trends over time without opening the browser UI. This is especially useful for team leads tracking how well AI-generated code is being reviewed across a sprint or time period.
Scope
cmd/stats.go(or similar)Beta Was this translation helpful? Give feedback.
All reactions