Records scope, branching, architecture, deletion gate, verification
protocol, path/config edits, migration ordering, and post-implementation
verification. Frames CI integration, scenario co-location, and Python
package rename as deferred work.
Per-file deletion of bash tests under superpowers/tests/ is gated by a
subagent that compares each bash assertion to its drill scenario's
verify block. Default keeps the bash test if any assertion is unmatched.
Branching: independent off dev (f/evals-lift), not stacked on
f/cross-platform.
Most new-harness PRs ship integrations that copy skill files or wrap
with `npx skills` instead of loading the using-superpowers bootstrap at
session start. Those integrations look like they work but skills never
auto-trigger.
Add an acceptance test ("Let's make a react todo list" must auto-trigger
brainstorming in a clean session) and require the transcript in the PR.