[WaterrrForever]

Follow-up to #7488, from the review nit there.

shadowingInstall named any older hyperframes on PATH and suggested npm install -g for anything it couldn't classify. A project-local node_modules/.bin/hyperframes also shadows plain npx hyperframes, but a global install never moves it.

Now the shadowing copy is classified the same way as the running one (runningInstall and installAt share the code):

• Global (npm / pnpm / bun / brew): the agent gets that manager's command, as before.
• Project-local: the agent gets the project's own manager, picked from its lockfile (bun add hyperframes@<v>, pnpm add …, yarn add …, npm install …). This reuses updateCommandFor, which doctor and upgrade already use.
• Unknown layout or workspace link: nothing is named. The skill's _meta.updateAvailable rule still pins the run.

The command now names the release the npx copy is running instead of latest, so a project install's command doesn't fall back to npx hyperframes@latest.

Tests: two new shadowingInstall cases. A project-local install with bun.lock gets bun add; an unknown layout gets null.

[jrusso1020]

Reviewed at 5403738d2e44aabc6cedefcd98d807c86fe5a24f.

The #7488 Nit is resolved. installAt classifies the shadowing path itself, while runningInstall retains its plugin override and prior classification order. Project-local candidates reuse the lockfile-aware updateCommandFor; unknown/workspace layouts no longer receive a guessed global command. The update targets the resolved run's version rather than a moving latest.

Validation: 180 tests passed across installerDetection, attendedTerminal, auth/login, updateCheck and autoUpdate. Replayed the earlier real-filesystem symlink repro against base and head: base suggests npm install -g hyperframes@latest; head correctly suggests npm install hyperframes@0.9.29. The new Bun-lockfile and unknown-layout tests also pass. Independent adversarial source pass found no new defect. No package installation/update was executed by these probes.

Need: yes · Size: ok · Reuse: shared installAt and updateCommandFor · Simpler: already minimal · Structure: fine (checked running-install and reverse consumers; no split)
Risk: low - reuses existing classification and command generation for a read-only notice.

Verdict: APPROVE
Reasoning: The follow-up closes the misleading project-local update suggestion without changing the existing installer behavior.

Reviewed by hf-review.

[github-actions]

Edit accuracy: accurate 267 (base branch 267), smooth 97 of those

The gate passes.
Smoothness is reported in the artifact, not gated. A case fails only if it fails 2 of 3 runs.

Quarantined, measured but not gated (8)

• seqplainkeys-none-center-r0-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-center-r0-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): fail / pass / pass, passes
• seqplainkeys-none-center-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-pct-r0-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-pct-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-px-r0-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-px-r30-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-px-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run