1145 narrowed local-release to install-only (binary + codesign-heal
wrapper) and removed the cargo build + gateway hot-restart steps that
the script used to do. That broke the "rebuild the gateway" muscle
memory: running script/local-release no longer rebuilt or restarted
anything, just re-installed the same binary.
Restore the build + restart logic while keeping 1145's wrapper:
- `cargo build --release --bin huskies` before install
- Snapshot the prior binary to ~/bin/huskies-bin.prev for rollback
- Print PREV → NEW version delta after install
- Detect a running `huskies .*--gateway` process and SSH-safe-restart
it (kill descendants depth-first, then nohup the wrapper from the
detached subshell)
- Wait up to 10s for the new gateway PID to appear; on timeout, roll
back to the previous binary and try to relaunch it
- Refuse to restart when more than one --gateway process matches, so
we don't kill the wrong tree
- `--skip-check` bypasses script/check for already-verified changes
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two issues that surfaced when story 1 ran in the adopted huskies-server
sled:
1. Dockerfile.base: the base image had no nodejs / claude CLI, so every
coder agent spawn in an adopted project sled failed with
`Unable to spawn claude: No viable candidates found in PATH`. Install
nodejs + @anthropic-ai/claude-code in the base image so every sled
built from it can spawn agents out of the box.
2. worktree/create.rs::install_pre_commit_hook: `git config --worktree`
requires `extensions.worktreeConfig = true` to be set on the repo
config; without it, every worktree creation logged a noisy
`Pre-commit hook install failed` warning. Enable the extension
idempotently before the per-worktree hooks-path set so the hook
install succeeds cleanly.
After this, rebuild huskies-project-base and recreate any adopted
project containers to pick up the CLI.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Story 1130 added HUSKIES_HOST=0.0.0.0 so the server INSIDE a project
container binds to all interfaces, but the host-side `docker -p`
mapping was still `127.0.0.1:{port}:3001` and `127.0.0.1:{ssh_port}:22`
— reachable from the docker host only, blocking remote MCP clients
and out-of-host SSH onto the project container.
Switch host-side mapping to 0.0.0.0 for both the MCP and SSH ports so
project containers spawned via `new project` are reachable from
anywhere that can route to the docker host. Existing containers
created before this commit retain their localhost-only mapping and
need to be recreated to pick up the change.
Add a regression test asserting both -p arguments use 0.0.0.0 and
reject any 127.0.0.1 restriction in the mapping.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Timmy
dave
