chore: discourage LLM-generated PRs (#26211)

* chore: discourage LLM-generated PRs

* chore: add reading CONTRIBUTING.md to PR checklist

* chore: workflow to close LLM-generated PRs

.github/pull_request_template.md

@@ -26,6 +26,7 @@ The `/api/something` endpoint is now `/api/something-else`
 
 ## Checklist:
 
+- [ ] I have carefully read CONTRIBUTING.md
 - [ ] I have performed a self-review of my own code
 - [ ] I have made corresponding changes to the documentation if applicable
 - [ ] I have no unrelated changes in the PR.

.github/workflows/close-llm-pr.yml

@@ -0,0 +1,38 @@
+name: Close LLM-generated PRs
+
+on:
+  pull_request:
+    types: [labeled]
+
+permissions: {}
+
+jobs:
+  comment_and_close:
+    runs-on: ubuntu-latest
+    if: ${{ github.event.label.name == 'llm-generated' }}
+    permissions:
+      pull-requests: write
+    steps:
+      - name: Comment and close
+        env:
+          GH_TOKEN: ${{ github.token }}
+          NODE_ID: ${{ github.event.pull_request.node_id }}
+        run: |
+          gh api graphql \
+            -f prId="$NODE_ID" \
+            -f body="Thank you for your interest in contributing to Immich! Unfortunately this PR looks like it was generated using an LLM. As noted in our CONTRIBUTING.md, we request that you don't use LLMs to generate PRs as those are not a good use of maintainer time." \
+            -f query='
+              mutation CommentAndClosePR($prId: ID!, $body: String!) {
+                addComment(input: {
+                  subjectId: $prId,
+                  body: $body
+                }) {
+                  __typename
+                }
+
+                closePullRequest(input: {
+                  pullRequestId: $prId
+                }) {
+                  __typename
+                }
+              }'

CONTRIBUTING.md

@@ -17,7 +17,7 @@ If you are looking for something to work on, there are discussions and issues wi
 
 ## Use of generative AI
 
-We generally discourage PRs entirely generated by an LLM. For any part generated by an LLM, please put extra effort into your self-review. By using generative AI without proper self-review, the time you save ends up being more work we need to put in for proper reviews and code cleanup. Please keep that in mind when submitting code by an LLM. Clearly state the use of LLMs/(generative) AI in your pull request as requested by the template.
+We ask you not to open PRs generated with an LLM. We find that code generated like this tends to need a large amount of back-and-forth, which is a very inefficient use of our time. If we want LLM-generated code, it's much faster for us to use an LLM ourselves than to go through an intermediary via a pull request.
 
 ## Feature freezes