← ClaudeAtlas

sync-envslisted

Read-only status/preview for claude-sync-by-skill — show what would change between this machine and the shared sync folder in both directions, without modifying anything. Use when the user types /sync-envs or asks to preview/check their Claude environment sync status. To actually move files, use /sync-env-up (push local → folder) or /sync-env-down (pull folder → local). For first-time configuration, run the engine with --setup.
JonMcMillan/claude-sync-by-skill · ★ 0 · AI & Automation · score 72
Install: claude install-skill JonMcMillan/claude-sync-by-skill
# sync-envs — status / preview (read-only) This skill **previews** the unified-environment sync without changing anything. The engine is `sync_engine.py` in this skill's folder. It never prints secret file contents. ## How to run The engine lives next to this file. Invoke it with the platform's Python 3: - **Windows:** `py -3 "<thisSkillDir>\sync_engine.py" --status` - **macOS/Linux:** `python3 "<thisSkillDir>/sync_engine.py" --status` (A launcher is also provided: `run.cmd` on Windows, `run.sh` on POSIX — e.g. `run.sh --status`.) `--status` is the default, so running with no arguments does the same. **Always launch it in the background** (the Bash tool's `run_in_background: true`). The sync folder is usually a cloud drive, and when the drive client is unwell the engine blocks in a kernel filesystem read for minutes; a foreground run holds the whole turn hostage so the user cannot ask you to investigate while it hangs. Backgrounding costs nothing on a fast run and keeps you reachable on a slow one. **The 30-second check (do this every run).** A healthy scan reaches its first output within a few seconds, so if the run has not finished at ~30s, check on it rather than waiting. The engine heartbeats to **stderr** while scanning (`[scan] claude-memory (up): 1840 files`); `--no-progress` suppresses it. Read the background output at ~30s and again ~10s later: if the counter is advancing it is healthy, just large; if the last `[scan]` line is identical across both samples (or