rch

# RCH — Remote Compilation Helper

`rch` transparently offloads compilation commands to remote workers via a Claude Code PreToolUse hook. The daemon picks the fastest healthy worker, rsync's the workspace, runs the build, syncs artifacts back, and exits with the worker's exit code.

This skill is the operational layer agents use when something about that pipeline isn't working — and the much more common case where it *thinks* it's working but is silently falling back to local execution. The skill is built around a single principle: **self-resolve before asking the human.** Every recovery path here is one the agent can run on its own.

Tested against rch v1.0.18; concepts apply to v1.0.16+.

---

## Read This First

When a build feels slow, run **one** thing:

```bash
RCH_VISIBILITY=verbose <your-command> 2>&1 | grep -E '^\[RCH\]'
```

The summary line is a contract:

| Pattern | What to do |
|---|---|
| `[RCH] remote <worker> (...)` | Healthy. Done. |
| `[RCH] remote <worker> failed [RCH-Exxx] ...` | Real build/env failure. See `references/ERROR_CODES.md`. |
| `[RCH] local (<reason>)` | **Fail-open.** See `references/FAIL_OPEN.md` and look up the reason verbatim. |
| *no `[RCH]` line at all* | Hook didn't fire. Run `scripts/protocol_test.sh "<your-command>"`. |

If you can't see why offload isn't happening, **prove the path works in isolation** before doing anything else:

```bash
rch exec -- env CARGO_TARGET_DIR="${TMPDIR:-/tmp}/rch_target_$(basename "$PWD")" cargo check --workspace --all-targets
```

If that prints `[RCH] remote <worker> (...)`, the offload pipeline is healthy. The problem is upstream of `rch exec` — usually the hook classifier or the agent's invocation form. If it also fails, follow `references/RECOVERY_PLAYBOOKS.md`.

---

## Quick Start

```bash
rch check                                                  # quick health
rch status --workers --jobs                                # full status
rch workers probe --all                                    # connectivity
rch hook status                                            # hook installed?
rch agents status                                          # which agents are wrapped
rch diagnose --dry-run "cargo check --workspace --all-targets"
rch exec -- env CARGO_TARGET_DIR="${TMPDIR:-/tmp}/rch_target_$(basename "$PWD")" cargo check --workspace --all-targets
rch self-test --all                                        # end-to-end verify
rch doctor                                                 # comprehensive checks
rch doctor --fix --dry-run                                 # see what doctor would auto-fix
```

If `rch status` shows storage pressure, always check both `/` and `/tmp` on the worker before deciding what to fix:

```bash
ssh ubuntu@<host> 'df -h / /tmp && free -h && cat /proc/pressure/memory && cat /proc/pressure/io'
```

---

## Fast Triage Order

Run in this order and stop at the first failing stage:

1. **Availability**
```bash
rch check
rch status --workers --jobs
rch workers probe --all
rch queue
```

2. **Config + socket consistency**
```bash
rch config show --sources
rch --json config get general.socket_path
rch --json daemon status
```

3. **Hook integration**
```bash
rch hook status
rch agents status
rch hook install                       # idempotent
```

4. **Command classification + path closure**
```bash
rch diagnose "cargo build --release"
rch diagnose --dry-run "cargo test --workspace"
```

5. **Remote compile proof**
```bash
rch exec -- env CARGO_TARGET_DIR="${TMPDIR:-/tmp}/rch_target_<name>" cargo check --workspace --all-targets
```

6. **If sync fails or storage looks bad, inspect the worker directly**
```bash
ssh ubuntu@<host> 'df -h / /tmp'
ssh ubuntu@<host> 'du -sh /tmp/rch-* /tmp/rch_target_* 2>/dev/null | sort -h'
ssh ubuntu@<host> 'find /data/projects -maxdepth 2 -type d \( -name "target_rch_*" -o -name "target_*" -o -name "target-*" -o -name target \) -exec du -sh {} + 2>/dev/null | sort -h | tail'
```

---

## Quick Fixes

| Symptom | Command |
|---------|---------|
| Hook not installed | `rch hook install && rch hook status` |
| Daemon not running | `rch daemon start` |
| Daemon version drift after upgrade | `rch daemon restart -y` (it drains gracefully — safe by default) |
| Socket mismatch / stale daemon state | `rch daemon restart -y` then `rch --json daemon status` |
| No workers configured | `rch workers discover --add --yes && rch workers setup --all` |
| Workers unreachable | `rch workers probe --all` then fix SSH key/host reachability — or `references/SSH_KEY_RECOVERY.md` if keys are missing |
| All workers busy | Queueing is on by default; if you still see fail-open, bump `RCH_DAEMON_WAIT_RESPONSE_TIMEOUT_SECS=120 <your-command>` (or raise `total_slots`) |
| Need a per-command priority bump | `RCH_PRIORITY=high <your-command>` (low\|normal\|high) |
| Want CARGO_TARGET_DIR / RUSTFLAGS forwarded to remote | `RCH_ENV_ALLOWLIST=CARGO_TARGET_DIR,RUSTFLAGS <cmd>` |
| Transfer churn under target dirs | Add excludes in `~/.config/rch/config.toml`, then `rch daemon reload` |
| Path dependency missing remotely | See `references/PATH_DEPENDENCIES.md` (now configurable via `[path_topology]`) |
| Sync fails with `Permission denied` in `/data/projects/<repo>` | Fix remote mirror ownership: `ssh ubuntu@<host> 'sudo chown -R ubuntu:ubuntu /data/projects/<repo> && sudo chmod 775 /data/projects/<repo>'` |
| Worker shows disk pressure (RCH-E210/211/...) | `references/DISK_AND_PRESSURE.md` — hand off to the `sbh` skill |
| Telemetry / SpeedScore broken | `references/TELEMETRY_RECOVERY.md` — move db aside, restart |
| Hook says installed but isn't intercepting | `scripts/protocol_test.sh "<your-command>"` |
| Worker shows pressure warning | Check `/` and `/tmp` separately, then inspect stale `rch_target_*`, `rch-*`, and `target_rch_*` dirs before broader cleanup |
| Multiple agents racing on fleet ops | Wrap with `scripts/multi_agent_safety.sh <cmd>` and use Agent Mail file reservations |
| Need