dough-story-wrap-up · diff

git:20260911.c7f6165 to git:20260912.5ace9a1

58 added, 114 removed. Audit A to A.

---

name: dough-story-wrap-up

description: >-

Closes one completed feature story or bounded retrospective correction using

available execution context. Assimilates lasting product knowledge, handles

existing follow-ups and authorized product decisions, removes that work's spent

plan and history so Git can recover it, and reports truthfully when required

inputs are missing or unfinished. Use to wrap up a story or correction, close

completed work, or delete spent plan and execution history.

---

# Story wrap-up

Close one selected feature story or bounded retrospective correction when the

coordinator invokes wrap-up. Use available execution context and optional

retrospective advice. Leave this project with maintained product knowledge and

no spent source or plan history in the current snapshot. Do not invent

findings, records, or a requirement for another conversation.

## Resolve this project's context

Require one selected work identity. Use a canonical feature story only when it

is explicitly supplied as that work's active home. Otherwise use the bounded

correction plan itself when it satisfies the correction-input contract defined by

[planning scope and lifecycle](../dough-story-refinement/references/planning.md#choose-the-planning-level).

Do not require or create a seed for that correction. Name any missing required

field, leave the affected work intact, and stop before closure.

Resolve from this project, not this skill's location:

- repository root and Git working tree;

- canonical seed location, story identity, heading or stable-anchor conventions

for a selected feature story;

- for planned work: executable-plan location, status vocabulary, and the

selected work's plan identity;

- for planless feature work: the story, its changes, and available execution

results;

- optional retrospective advice when it is present, including an empty

result;

- Git commit conventions used to preserve a recoverable revision;

- for planned execution, its selected mode and available originating checkout,

execution checkout and branch, and integration-target identity;

- the product backlog path when a **Taken** or **Backlog list** entry points

at the selected work; and

- shared records that name the selected work: its seed when applicable, process

log (`DearDough.md` unless this project sets another canonical location),

incoming links, and assessment or recognition records.

Do not invent a retrospective artifact or require another review. Do not invent

a plan path, completion rule, feature-story seed location, or Git convention.

Planless feature work does not require a plan. Keep the bounded-correction

identity contract above; do not invent a planless correction format.

If context needed for a closure decision is missing, name the gap, leave

affected material intact, and do not claim closure.

Before deleting a plan that carries planned-execution identity, retain the

resolved mode and checkout, branch, and target values in the coordinator's

available execution context for the remaining wrap-up actions and report. Do

not create a parallel registry. Missing identity needed by a later action stops

that action instead of reconstructing or guessing it after plan deletion.

## Establish execution completion

Judge completion from the selected work and available execution evidence.

- Planned work is complete when every slice is done.

- Planless feature work is complete when the supplied story, changes, and

execution results show the promised outcome is delivered.

- Incomplete implementation leaves the affected active work intact. Report

the unfinished implementation and stop. Do not delete an unfinished

correction plan to manufacture a wrap-up.

When retrospective advice is present, apply it under existing authority in the

closure actions below. When it is absent or empty, perform the same ordinary

closure using the other available inputs: the story or bounded-work context,

implementation results, maintained product knowledge, existing follow-ups, and

coordinator instructions.

## Assimilate lasting knowledge

Move lasting behavior and design into this project's maintained code, tests,

documentation, or current Accepted decisions. Describe the current product

without execution narration, impact chronology, story or plan identity, or

retrospective judgments. Preserve existing product tests and documents that

already state current behavior. Do not invent product knowledge. When a current

product fact is written only in spent execution context or review and is not

already stated in maintained documentation, write it into maintained

documentation before deleting that spent copy. Tests that exercise related

behavior do not replace that documentation step.

For North Star topics cited, added, or revised by the completed work, apply the

shared [topic-retirement

instructions](../dough-slice-planning/references/architectural-thinking.md#retire-temporary-direction-during-ordinary-wrap-up)

as part of ordinary closure.

## Queue an existing follow-up plan first

When an existing follow-up plan is present, validate it against the

correction-input contract above, then put its one canonical active home first

in the queue. Do not refine, replan, or execute it. Preserve the plan contents

needed for later execution. Handle that follow-up by its presence, including

when retrospective advice is absent.

Resolve one canonical active home under

[dough-product-backlog](../dough-product-backlog/SKILL.md#canonical-active-homes):

- If a canonical follow-up story is supplied, keep that story in its seed and

link the plan there. Queue the story; do not duplicate it as a plan entry.

- Otherwise queue the existing correction plan directly. The plan is its

canonical active home; do not create or recover a seed solely for queueing.

- If required correction input is missing, name the missing field and do not

guess or queue the addition. Keep the follow-up plan and any other needed

active-work context, report the gap, and stop before deleting the completed

predecessor's history.

Preserve unrelated queue order after that first item, near-future direction,

and human text. Repeating wrap-up must recognize either canonical home and must

not duplicate the follow-up or queue entry.

## Apply product-review decisions

When retrospective product advice or additional human input is present, apply

only authorized compatible backlog and canonical-home changes. Follow

[dough-product-backlog](../dough-product-backlog/SKILL.md) for queue and

active-home conventions. Explicit human input wins over advice.

Supported changes: relevant reorder, queue membership, understood new-story

addition, understood bounded-correction plan addition, and canonical-detail

edits. Apply the backlog skill's canonical-home admission rules to every new

entry. Keep existing follow-up work first unless a later explicit human

instruction changes that priority. Preserve unrelated content, still-needed

acceptance work, and near-future direction.

A skipped, empty, or absent product review, or absent extra human input,

introduces no mandatory question. Ordinary closure continues using the other

available inputs. Leave unresolved necessary context with active work and

report the choice; do not invent scope, launch discovery, or start another

review.

## Leave no extra wrap-up ceremony

Do not launch discovery or another review. Absent or empty retrospective

advice is valid and changes nothing beyond the supported closure and follow-up

actions above.

## Commit closure inputs and preserve Git recovery

After supported follow-up queue changes and before deleting anything, make the

current revision recoverable with this project's ordinary Git conventions.

Commit all owned review and closure-input changes in that revision, including

applicable retrospective edits to the process log, an uncommitted active

follow-up plan and its queue edit, assimilated product knowledge, and the spent

material. Preserve unrelated changes and include only files or portions whose

ownership is unambiguous. Resolve ownership of the intended cleanup targets at

this boundary too, before deleting any of them. Then record that revision as

the before-cleanup commit, even when the current revision was already suitable.

If commit conventions, ownership, or recovery cannot be resolved, leave the

material intact, report the gap, and do not claim closure.

Preserve Git history as the sole recovery surface for spent execution material.

Keep current product knowledge in maintained project content.

## Delete spent history, including shared records

- When completion and recovery are resolved, delete only material the selected

- work's identity and references identify as spent:

- - its executable plan and owned proof, evidence, and assessment records, when

- a plan exists. A plan decision that retained the plan at execution completion

- still leaves that spent plan for wrap-up to delete.

- - its canonical story section when it has one; delete the seed only when every

- remaining section is spent;

- - the completed work's entry in **Taken** or **Backlog list**;

- - related occurrences in the process log, and an issue or container that

- becomes empty afterward; and

- - incoming links that exist solely to preserve that spent history.

+ After completion and Git recovery are established, delete the selected work's

+ spent history:

- After deleting spent files, remove directories named by that spent work when

- they are empty, including nested untracked evidence directories. Verify those

- directory paths are absent, not merely free of files.

+ - its executable plan and owned proof, evidence, and assessment records, even

+ when the plan was retained at execution completion;

+ - its canonical story section, and its seed only when every remaining section

+ is spent;

+ - its **Taken** or **Backlog list** entry;

+ - its process-log occurrences, and issues or containers left empty; and

+ - links whose sole purpose is preserving that history.

- Preserve unrelated human text, sibling stories, unrelated log issues and

- occurrences, product and version identity, maintained tests or documents,

- still-needed acceptance work, and any follow-up plan queued above. A direct

- queue link to that plan is active navigation, not an incoming historical link

- to delete with the completed predecessor. Preserve the active plan's correction

- input and provenance. When its source locator points into the predecessor being

- deleted, replace that locator with the before-cleanup commit and the predecessor's

- repository-relative path rather than deleting the active context or recreating

- spent history.

- Resolve ambiguous attribution before deleting that portion; if a log issue or

- link cannot be tied to the selected execution, leave it intact and say so.

+ Remove empty directories belonging to the spent work, including untracked

+ ones. The current snapshot must be free of that history, both tracked and

+ untracked, with recovery available from the recorded before-cleanup commit.

- Repair remaining Markdown links that this deletion breaks; do not leave a

- live link to a removed path. Unrelated links stay unchanged.

+ Preserve unrelated human text, sibling stories and log entries, product and

+ version identity, maintained tests and documents, still-needed acceptance work,

+ and active follow-ups. Shared records lose only the portions attributable to

+ the completed work; leave uncertain portions intact and report the ambiguity.

- Keep maintained content focused on current product knowledge, with the removed

- work's identity and execution judgments recoverable through Git. An already-absent

- artifact does not prove different work complete. Repeating wrap-up must not recreate

- history, duplicate edits, or claim that missing files close a different story

- or correction.

+ Keep an active follow-up's queue link, correction input, and provenance. If its

+ source locator points into deleted predecessor history, replace the locator

+ with the before-cleanup commit and repository-relative path. Repair Markdown

+ links broken by cleanup without recreating spent history.

- Inspect tracked and untracked files. Absence is the current snapshot, including

- untracked paths. Recover removed files with

- `git show <before-cleanup-commit>:<spent-path>` using the recorded revision.

+ Repeating wrap-up must recognize already-completed cleanup without duplicating

+ edits. Missing artifacts alone do not establish completion of another work item.

## Commit final closure

- After the spent-history deletion and link repair above, inspect the complete

- closure diff and commit all owned closure changes with this project's ordinary

- Git conventions. Preserve unrelated staged and unstaged changes. A successful

- wrap-up requires a committed final snapshot; staged or unstaged deletion is not

- completion. If ownership or commit completion is ambiguous or the commit fails,

- retain the material and Git state, report the unresolved closure, and do not

- claim success.

+ Review and commit the owned closure changes using this project's Git conventions,

+ preserving unrelated changes. Both the before-cleanup revision and the final

+ closure must be committed in either execution mode; uncommitted cleanup is not

+ completion. Report unresolved ownership or commit failures without claiming

+ closure.

- Finish both the before-cleanup and final-closure commits in caller-selected

- direct-current-branch mode as well as worktree mode. In worktree mode, complete

- these commits before any later integration or owned worktree/branch removal.

- Direct-current-branch mode has no later integration or worktree-removal action.

+ Direct-current-branch mode ends with committed closure. Worktree mode continues

+ with integration and resource cleanup below.

## Integrate committed worktree closure

- For worktree mode, save the committed final-closure tip and use the retained

- planned-execution identity to resolve the exact integration-target branch and

- the checkout that holds it. Leave the execution checkout before integration;

- run target inspection and integration from the target checkout. If the target

- checkout cannot be resolved uniquely, does not hold the recorded target branch,

- or no longer matches the retained identity, preserve the execution branch and

- worktree, report the mismatch, and stop this action. Direct-current-branch mode

- skips integration.

- Inspect the target checkout's branch, staged and unstaged changes, untracked

- files, and any unfinished Git operation before merging. Preserve unrelated

- target work. Do not stash, reset, overwrite, silently include it, or proceed

- through a state whose safety or ownership is ambiguous. An unsafe target state

- stops integration with both the execution branch and worktree intact.

+ Save the committed final-closure tip and integrate it into the recorded target

+ branch from its checkout, using the retained execution identity and this

+ project's ordinary local merge conventions. Preserve unrelated target work. If target identity or

+ safe integration cannot be established, retain the execution resources and

+ report the blocker. Do not rebase, push, delete remote branches, or add CI waiting.

- First test whether the saved execution tip is already an ancestor of the target

- branch. If so, treat integration as already done and do not merge again.

- Otherwise merge that committed tip into the target branch using this project's

- ordinary Git merge conventions. This integration is local: do not rebase, push

- the target branch, delete a remote branch, or introduce CI waiting policy. A

- merge conflict remains in the target checkout for explicit resolution; do not

- abort, reset, remove the execution worktree, or delete either branch. Report

- the conflicted paths and actual Git state, and do not claim wrap-up complete.

+ Resolve merge conflicts by understanding the reasoning behind both sides and

+ reconciling their intended behavior, using the surrounding code, history, and

+ available work context. Verify the resolution with appropriate checks and

+ complete the merge. Stop conflict resolution only when the available evidence

+ cannot justify a coherent resolution; explain the incompatible intentions or

+ missing decision, report the conflicted paths and Git state, and preserve the

+ work for a human decision.

- After a successful merge or an already-integrated result, verify that the saved

- execution tip is an ancestor of the recorded target branch. A failed ancestry

- check is unresolved integration: preserve the branch and worktree, report the

- observed refs, and do not claim completion. Successful integration alone does

- not remove those owned resources; keep them for the later safe cleanup action.

+ Integration is complete when the saved final-closure tip is an ancestor of the

+ recorded target branch and the target contains the committed closure. Recognize

+ an already-integrated tip without merging again. Unresolved integration leaves

+ the execution branch and worktree intact and blocks wrap-up completion.

## Remove integrated worktree resources safely

- After verified worktree-mode integration, use the retained execution identity

- to inspect the current state of the exact execution-checkout path and local

- execution branch. For each resource still present, confirm from Git's worktree

- and ref state that its path, checked-out branch, and tip match the retained

- identity, and that the path is not the originating or target checkout. Treat an

- identified resource's absence as an already-completed cleanup step only when

- the current Git state contains no conflicting resource at that identity. A

- missing retained identity, changed present resource, or ambiguity stops cleanup;

- do not infer ownership from a branch name or reconstruct it from history.

- When the execution checkout is present, inspect its tracked changes, untracked

- files, index, and unfinished Git-operation state. Recheck that the saved

- execution tip is an ancestor of the recorded target branch whether resources

- are present or already absent.

- Dirty tracked or staged changes, untracked content, an unfinished operation, or

- another ownership ambiguity leaves the execution worktree and branch intact.

- Report the retained data and the already-completed integration separately. Do

- not stash, reset, clean, or force removal.

- From a surviving checkout outside the execution directory, remove the exact

- owned clean worktree when it remains with ordinary non-force `git worktree

- remove`, then delete the exact integrated local execution branch when it remains

- with ordinary non-force `git branch -d`. Never remove the originating checkout

- or a caller-owned checkout or branch. Never force either operation or delete a

- remote branch.

+ After verified integration, remove only the clean execution worktree and local

+ execution branch identified by the retained execution context. Confirm their

+ current identity and that no unique work, including untracked content or an

+ unfinished Git operation, would be lost. Preserve originating and target

+ checkouts and caller-owned resources. Missing or changed identity, dirty state,

+ or uncertain ownership blocks the affected cleanup; report what remains.

- Treat each cleanup operation independently. If worktree removal succeeds but

- branch deletion fails, retain the branch and any other reported resource; do

- not escalate to force or remove an unrelated worktree that now uses it. Report

- integration as complete but cleanup as partial, without the completion marker.

- On retry, use the retained saved identity and current Git worktree and ref state.

- Recognize already-absent owned resources without recreating history, and stop on

- an identity mismatch instead of selecting a similarly named branch or path.

+ Use non-force removal from a surviving checkout. Do not discard or stash work

+ to enable cleanup. Recognize already-removed owned resources on retry without

+ selecting substitutes. If cleanup only partly succeeds, retain the remaining

+ resources and report integration and partial cleanup separately.

- Report successful closure only from a surviving checkout after verifying all of

- these outcomes: the execution directory is absent, Git's worktree listing no

- longer contains its path, the local execution branch is absent, the saved

- execution tip remains an ancestor of the recorded target branch, and the target

- contains the committed closure. Direct-current-branch mode has no cleanup

- action and never treats its caller-owned checkout or branch as spent execution

- resources.

+ Completion requires the execution directory, its Git worktree registration,

+ and its local branch to be absent, with the committed closure still integrated

+ in the recorded target. Direct-current-branch mode has no resource cleanup.

## Report

Report the selected work and its canonical identity, completion judgment,

execution mode and retained checkout/branch/target identity when applicable,

before-cleanup and final-closure commits when deletion happened, assimilated

knowledge, deleted paths, the saved execution tip and local integration result

in worktree mode, worktree and local-branch cleanup results, preserved

unsupported material and resources, and any gap that blocked closure.

Distinguish a new merge from an already-integrated tip, integration success from

partial or refused cleanup, and name unsafe or conflicted state without implying

that the target was pushed or a remote branch was deleted.

Distinguish a completed wrap-up from a refusal that left files intact.

End a successful closure with:

`## STORY WRAP-UP COMPLETE`

Do not emit that marker when required context, unfinished work, unresolved

recovery or worktree-mode integration, or remaining required worktree cleanup

blocks closure.