Repository navigation
foreman-start: progress in the workspace name, mechanical steps in a script - #2
Merged
Merged
Conversation
The foreman renames its own workspace to "Foreman (<done>/<total>)" once the plan is written, and updates it after every ticked item, added item and docs-only item. The count lives in the foreman block so a resume continues it.
One stdlib Python script now does what the model did by hand: find the TODO file, write the foreman block and checklist markers, keep the progress count and the "Foreman (done/total)" workspace name, move the item's own ticket, scan every worker for a heartbeat (with NO CHANGE when nothing needs the foreman), verify a PR is merged, and check the machine before raising the parallel limit. SKILL.md calls it instead of describing those steps. The .claude and .agents copies are now byte-identical: bin/sync-skills copies one to the other and its --check runs in CI.
…a file The foreman can work from a TODO file or from Jira tickets, so the count of items picked up in a run doesn't belong in either. Drop the progress line and the state init/total commands; the foreman keeps done/total itself and sets the workspace name with the new `name DONE TOTAL`. `name` without numbers reads the count back from the workspace name after a lost context. `check` no longer keeps a snapshot in ~/.cache/foreman either; it reports the current state only.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Foreman (done/total)and updates it after every finished item. It keeps that count in its own session only: the list can be a TODO file or Jira tickets, so neither stores it.scripts/foreman.py(stdlib Python) does the mechanical steps the model did by hand: pick the TODO file, write the foreman block and checklist markers, set the workspace name, move the item's own ticket (jira-cli or acli), scan every worker for a heartbeat (NO CHANGEwhen nothing needs the foreman), verify a PR is really merged, and check the machine before raising the parallel limit. SKILL.md calls it instead of describing those steps..claudeand.agentscopies of the skills are now byte-identical. Edit.claude/skills/and runbin/sync-skills; CI runsbin/sync-skills --check.Tested read-only against a live foreman run and its workers; everything that renames or moves tickets was tested with fake
paseo/jira/ghon PATH. Theaclipath is untested (not installed here).