Repository navigation
chore: extend pre commit instructions for AI agents - #23313
Conversation
|
|
||
| See the [Testing Quick Start](docs/source/contributor-guide/testing.md#testing-quick-start) | ||
| for the recommended pre-PR test commands. | ||
| If documentation files changed then run |
There was a problem hiding this comment.
Other ones I use:
ci/scripts/rust_prettier.sh
ci/scripts/rust_clippy.sh
There was a problem hiding this comment.
I cant see rust_prettier 🤔
There was a problem hiding this comment.
Sorry I meant
ci/scripts/rust_fmt.sh
There was a problem hiding this comment.
but I see those commands are already covered above
|
Thanks @alamb for the review |
|
FWIW, I find this change causes coding agents to run the extended tests way too often, which slows down local iteration quite a bit. |
The idea was use more local resources instead of ASF, probably agents try commit too often? |
|
Maybe we can adjust and say we don't need to run extended tests locally? |
|
The better possible option is to ask agent to identify if changes require extended tests to run, we have everything for that in CI rules so the agent would most likely do the correct thing. Supposedly this reduces number of runs. Removing extended tests to run locally, can hit us on merge queue failures, I won't be considering this. But for experienced contributors they can just amend locally the skill to avoid excessive runs if they feel this is a problem |
Yes I agree, it's also easy to override default AI instructions locally. |
Which issue does this PR close?
Rationale for this change
Extend pre commit instructions for AI agents to involve contributor machine to find issues before being committed. Each commit triggers a CI which in its turn asks ASF Infra for their precious resources and if we can find and fix as early as possible, much before utilizing ASF resources, would be amazing
What changes are included in this PR?
Are these changes tested?
Are there any user-facing changes?