dsh-pstack 0.2.0-beta.1 → 0.2.0

package/assets/skills/poteto-mode/playbooks/visual-parity.md

@@ -1,11 +0,0 @@
-### Visual parity
-
-**You own pixel-exact equivalence. The baseline is the spec. You do not touch it.** Equivalence is verified by image diff, not by eye.
-
-1. Establish the baseline first, before any migration: a visual regression harness that screenshots the current component across its states, plus the target when matching two implementations. No baseline, no parity claim. A blocking prerequisite, not a follow-up. Freeze it before the first component moves: record the harness command and the baseline image paths, and do not regenerate a baseline mid-run, because a diff against a baseline you rewrote proves nothing.
-2. Anti-shortcut clauses, stated and held: no harness modifications, no baseline tampering, no component restructuring to make a diff pass. If the baseline looks wrong, stop and ask, don't edit it. Editing it is the one shortcut that makes every later diff meaningless.
-3. Migrate one component at a time. Parallelize across local worktrees, one owner per component, because two writers in one checkout overwrite each other (the **separate-before-serializing-shared-state** principle skill). Give each owner its own `create_worktree` checkout and pass that absolute path as its `cwd`; a worktree separates the working directory and the branch, not ports, caches, or services, so name whatever else the component shares. Shared primitives migrate first as a blocking phase.
-4. Verify each component against its baseline with a real image diff on the matching surface. Name the comparison tool and prove it separates before migrating anything. A nonzero diff is a fail, so investigate the pixel delta and iterate until it reaches zero. If this machine has no image comparison that reports a per-image delta, stop and say so: without it the run can only report what the eye saw, and that is not a parity claim. A goal can drive the iteration when the run is long, with a zero-diff predicate and `update_goal` `complete` only when every component's diff is zero, but a goal measures nothing itself, so a completed goal is not a parity verdict.
-5. Run **Opening a PR** per component or per safe batch.
-
-**Reply:** components migrated, the diff result for each, the comparison tool and the frozen baseline harness location, what's left, and whether parity is claimed or unverified.

package/assets/skills/poteto-mode/playbooks/worktree-cleanup.md

@@ -1,20 +0,0 @@
-### Worktree and simulator cleanup
-
-1. Snapshot and audit. Record `df -h /`, then run `pstack_script` with `id: "worktree-audit"` and `args: ["-b", "<pr-base-ref>", "--sessions", "<json>", "<repo-path>"]` (principle-build-the-lever). The tool runs the bundled script on the Host's own runtime, so no `bun`, `bash`, or `node` on the PATH is involved. Without `-b` it checks `MERGED` against `origin/HEAD`, so pass `-b origin/<branch>` when PRs target another branch. It reads paths from `git worktree list`, never hand-typed, since a hand-typed `myrepo-worktrees/x` misses one that lives elsewhere (principle-encode-lessons-in-structure). It classifies each worktree by size, age, merge state, uncommitted work, PR state, and the newest session that touched it, then suggests a bucket.
-
-    Its session input is an array you build from `pstack_history` results, one entry per session you read:
-
-    ```json
-    [{"sessionId": "<session.id>", "cwd": "<session.cwd>", "lastEventTime": 0}]
-    ```
-
-    Use the largest `events[].time` in that session's page as `lastEventTime`. The script never reads the session store, so an entry you omit is a worktree it cannot see, and omitting the whole array leaves `LAST_CHAT` as `-` and the bucket as `unknown-usage`. A `pstack_history` result passed through unchanged is refused: that tool returns one object, never an array. Pass `background: true` for a large tree.
-2. The bucket is advice, not permission. The pinned and active sessions are the real artifact (principle-prove-it-works). Pinned state has no public source in this host, so the script prints `PINNED unknown` for every row; ask the user for the pinned and in-use set instead of inferring it, and cross-check every candidate. When the user names a pinned or in-use worktree, that set wins over any `safe` bucket.
-3. Verify usage before deleting. For every `verify-recent-chat` or `unknown-usage` row, or anything you doubt, ask the user which sessions are still live, and read the ones you may with `pstack_history` to see which worktree paths those events touch. Do not scan the session store for paths (principle-guard-the-context-window: session logs are bulk). An active session can start arena and repro trees in sibling worktrees, so a worktree can be in use even when its name never appears in the user's own summary. Pick each subagent's role per the Subagents section of `poteto-mode`; read its `SKILL.md` at its absolute resource path. Start from the level the role line names, raise it for hard work and lower it for routine work by passing `reasoningEffort`; a child that cannot call `set_reasoning_effort` keeps the level you gave it.
-4. Pause on irreversible loss. `wip:N` is N tracked uncommitted edits. Show the diff and get a decision first, since removing a clean worktree is recoverable from its branch but uncommitted work is gone. `scratch:N` is untracked throwaway, safe to drop, but name the files. Per Autonomy, clean and merged and not-in-use proceeds. `wip` and in-use pause.
-5. Prune the confirmed set. Re-read `git worktree list` and confirm each absolute path with the user immediately before touching it, and remove only paths approved in this run. Per path, `git worktree remove --force <path>`. If the dir survives on ignored build artifacts, `rm -rf` that same verified path, then `git worktree prune`. Branch refs survive, so no commits are lost. Confirm with `df -h /` and re-list.
-6. Simulators and other reclaimers. Simulators are usually the next-biggest win. `xcrun simctl --set testing delete all` (XCTestDevices clones), `xcrun simctl delete unavailable`, and `xcrun simctl runtime list` then `runtime delete <id>` for old runtimes. More when needed: Xcode `DerivedData` and `iOS DeviceSupport`, `~/Library/Application Support/Cursor` (`state.vscdb.backup`, and `snapshots/roots/<root>` where a `<root>` named for a folder you opened as a workspace balloons), package caches (pnpm, uv, brew, yarn). Clear only caches the user has not said to keep.
-
-This is the one playbook that deletes user state with no code review to catch a slip, so the gates above are the review.
-
-**Reply:** `df -h /` before and after with space reclaimed, the worktrees pruned, and a one-line reason for each held back (in use by which session, or which uncommitted work). State the pinned set as the user gave it, and name anything still unknown.