Repository navigation
fix(status): count work merged from another branch by its Task: line - #52
Open
ancelin-code wants to merge 1 commit into
Open
ancelin-code wants to merge 1 commit into
ancelin-code wants to merge 1 commit into
Conversation
A task delivered by a pull request whose head is not task/<id> stayed ready, so where.sh kept proposing it and follow.sh would launch a bench to redo it. The bulk merged query now also reads "Task:" lines from pull request bodies, which publish-bench.sh already writes and a person can add by hand. One line may name several task files. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What and why
A task delivered by a pull request from a branch other than
task/<id>stayedreadyforever.where.shkept proposing it, andfollow.shwould launch a bench to redo shipped work. This happened in the tiroirretrievalcut withdocs-001anddocs-002(feedback item 1).status.shnow also counts a merged pull request whose body has a lineTask: specs/tasks/<id>.md.publish-bench.shalready writes that line; for a pull request opened by hand, a person adds it to the body. One line can name several task files.How to review
etabli-status/scripts/status.sh: the bulk merged query now fetchesbodytoo. Its--jqkeeps only the lines that start withTask:, and anawkstep turns eachtasks/<id>.mdpath on them into atask/<id>head for the existingrecord. Theghoutput is captured before parsing, so a failed query still stops the script instead of reading as "nothing merged".--head task/<id>. Adding a body search per task would double the GitHub calls on every poll. The limit is documented.--excludeoption forfollow.shandlaunch-benches.shthat the feedback suggested. It would be a second, local-only way to say the same thing, lost on every restart.How to try it
In a project whose task was shipped from another branch, edit that PR's body to add
Task: specs/tasks/<id>.md, then runsh .agents/skills/etabli-status/scripts/status.sh: the task readsdone, and its dependents becomeready.The
--jqfilter was also checked against realghoutput and a CRLF body, since the test fake does not run jq.Does it earn its place?
Without it, the only fix was to run
follow.sh --no-launchand launch every task by hand. A mention of a task file anywhere else in a body, such as a decomposition PR listing the files it adds, does not count; a test covers that. Recorded inCHANGELOG.md.🤖 Generated with Claude Code