fix(studio): review round 2 - a conversion that fails after its commit recovers without a restart

- The convert job reloads the course on success and on failure, so learners
  get the intact import (rolled back) or the published studio course, never a
  course whose folder moved away.
- POST /convert on a course with an incomplete conversion (.converting-<id>,
  staging or retired folders) queues a job that recovers and finishes it
  instead of answering 409; the studio page shows "Conversion incomplete" with
  Retry for the owner.
- The 409 for pending import jobs names their kinds and ids; the dead withLock
  in git.ts is gone; the dot-folder discovery rule is documented in
  docs/operations.md and spec section 5.
- Route tests (SQLite and Postgres) inject a failed cross-filesystem copy, a
  checksum mismatch and a pre-commit failure.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

docs/superpowers/specs/2026-10-30-stcourses-design.md

@@ -181,6 +181,7 @@ Phase 0 (section 10) pins down the exact schema. These principles are fixed:
   - The two reference courses must load by adding only a manifest and front matter, without moving or rewriting their existing files.
   - Their existing tooling (test scripts, `tools/*.mjs`) must keep working.
 - **Derived data lives in a cache directory that can be rebuilt from the files:** search index, slide images and text, transcripts, video posters and AI embeddings.
+- **Discovery** (v1.1 note): the platform finds `stcourse.yaml` at depth 0–2 under each course directory and skips folders whose name starts with `.` (and `node_modules`). A course folder must therefore not be a dot-folder. The studio relies on this to hide an import folder while it is converted (`.converting-<id>`, `.converted-<id>.<n>`).
 
 ## 6. Accounts, roles and visibility