feat: weekly workflow to solve non-actionable positive reviews - #38
Draft
Bilb wants to merge 4 commits into
Draft
Conversation
5,253 app-store reviews sit unsolved, 4,812 of them still `new`, against 428 non-review unsolved tickets — so the backlog is 92% reviews nobody was ever going to action. Solving them was already being done by hand: 4,959 are already solved or closed. resolve_reviews.py searches `via:any_channel status:new`, re-checks each ticket locally against triage.is_store_review and triage.review_stars (the query cannot express the rating, and detection must not drift from the triage's), and PUTs solved in batches of 100 through update_many, polling each asynchronous job so per-ticket failures fail the run. It writes to Zendesk, so the guards are the design: nothing happens without --apply, a scheduled run applies but a manual dispatch is a dry run unless asked, `new` only by default since an `open` review may have had agent activity, `solved` never `closed`, `additional_tags` rather than `tags` so a ticket's own tags survive, and the note is private because a public comment would email the reviewer. Solving can fire triggers — including satisfaction surveys to real reviewers — which the script, workflow and README all flag, with a small first run recommended. 26 offline tests cover the guards rather than the happy path. Not run against Zendesk yet, by request.
"Backlog: 5,680 unsolved" read as 5,680 things wanting attention when 92% of them are AppFollow reviews and the human queue is 428. The line now leads with the review-excluded number and keeps the review count beside it, so the figure a reader acts on is the one that means something. A second count-only search does it — one extra request, no fetching, the same endpoint the existing line already uses. Both counts stay best effort: if the new one fails the line falls back to the plain total rather than vanishing, which a test pins.
The flag was a hedge, not a decision: I added it without knowing what an `open` review actually is, so it offered a widening nobody had a reason to use. Checking settles it — in a 100-ticket sample of the 441 open reviews, every single one had an assignee, a group, and an updated_at later than its created_at. Something has already handled those, which is exactly the class of ticket a bulk status change should leave alone. So `status:new` is now the query, not a default, and the reason is recorded next to it rather than a guess about agent activity. The workflow input and the flag plumbing go with it.
--min-stars let a dispatch lower the floor, and lowering it is exactly the wrong move: 3★ and below are the reviews the triage reads as bug reports in disguise, so a run at --min-stars 3 would close the ones most worth looking at. Renamed DEFAULT_MIN_STARS to MIN_STARS since it is no longer a default, and dropped the flag, its 1-5 validation, and the workflow input. select_resolvable still takes the floor as an argument — it stays a pure function the tests can drive at any threshold — but main passes the constant and nothing else can. Verified with a guarded dry run: 177 of 200 fetched would be solved, GET only.
Bilb
marked this pull request as draft
August 7, 2026 09:38
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Not started with the --apply just yet.
Probably worth not merging it before I can make sure it works correctly (gh needs the workflow to be on the main branch to run it).