Sweeps terminal-status planning entities into per-kind archive subdirectories as a single git commit.
npx skills add https://github.com/23min/aiwf --skill aiwf-archive-23min
Please help me install this Agent Skill. Skill: aiwf-archive Source: https://github.com/23min/aiwf/tree/main/internal/skills/embedded/aiwf-archive Command: npx skills add https://github.com/23min/aiwf --skill aiwf-archive-23min
aiwf archive verb to move terminal-status entities into per-kind archive/ subdirectories while preserving id-based references and git-history provenance. ## Core Features & Use Cases - Dry-run by default: Preview every planned move without touching the tree, then commit the sweep with a single --apply flag flip. - Drift control: Configure archive.sweep_threshold in aiwf.yaml to escalate the archive-sweep-pending finding from advisory warning to a push-blocking error. - Skip semantics: Entities with uncommitted changes or non-terminal children are skipped and named, never half-moved; the rest of the sweep still commits. - Use Case: After closing a milestone and finishing wrap rituals, run a dry-run to preview the pending sweep, confirm the counts, then apply it as one commit carrying the aiwf-verb: archive trailer. ## Quick Start Ask the AI to run a dry-run of aiwf archive, review the planned moves and skipped entities, then apply the sweep as a single git commit.High-intent search queries and answers about installing and using this skill.
Run `aiwf archive` to preview the planned moves as a dry run, then re-run with `aiwf archive --apply` to commit the sweep as a single git commit. You can scope it to one kind with `--kind gap` or another entity kind.
The dry run is the default behavior: running `aiwf archive` with no flags prints the planned moves and any skipped entities without modifying the tree. Re-run with `--apply` only after reviewing the output.
A candidate is skipped when a file the sweep decision depends on has uncommitted changes, or when an epic still has non-terminal child milestones. Commit or revert the named file, or close the child milestone, then re-run the sweep.
No reverse verb exists by design; the FSM is one-directional and archive is its structural projection. The canonical pattern is to file a new entity referencing the archived one, since id references resolve across both active and archive directories.
Set `archive.sweep_threshold` in aiwf.yaml to escalate the archive-sweep-pending finding from a warning to a push-blocking error when pending sweeps exceed the configured count. The default is unset, which keeps the finding advisory.
No. Milestones live as flat files inside their parent epic's directory and ride with the epic when it archives. A done milestone under an active epic stays in place until the whole epic subtree moves.
menkesuMap PM promotion paths and identify skill gaps with action templates.
JongDeugGroups changes into commits and drafts Conventional Commits messages with approval gates.
the-perfect-developerAutomate local Git workflow validation with pre-commit checks and push triggers.
endorphin-aiManage project documentation and work items using markdown files with Git.