feat(nanoclaw): switch to nanocoai upstream and pnpm-based install#33
Open
ealeyner wants to merge 4 commits into
Open
feat(nanoclaw): switch to nanocoai upstream and pnpm-based install#33ealeyner wants to merge 4 commits into
ealeyner wants to merge 4 commits into
Conversation
Track the upstream nanoclaw repo move from qwibitai/nanoclaw to nanocoai/nanoclaw, and rework the install to match the project's current toolchain: Node 22 + pnpm v10, with the OneCLI post-install check requiring ~/.local/bin on PATH. The kit now boots directly into the nanoclaw service and runs the setup wizard on attach, instead of dropping into Claude Code. - spec.yaml: clone from github.com/nanocoai/nanoclaw, install via pnpm with --network-concurrency=1, upgrade Node to 22, install pnpm globally, persist ~/.local/bin in PATH - spec.yaml: add allowed domains for the new install + setup paths (nodejs.org, onecli.sh, raw.githubusercontent.com, release-assets.githubusercontent.com, ghcr.io, registry-1.docker.io, auth.docker.io, production.cloudflare.docker.com, checkpoint.prisma.io) - spec.yaml: NODE_NO_WARNINGS=1, ONECLI_BIND_HOST=127.0.0.1 - spec.yaml: entrypoint starts nanoclaw service, waits for the CLI socket, then runs setup:auto with NANOCLAW_SKIP=service,container - spec.yaml: drop aiFilename/containerSettings.claude (kit no longer shells out to Claude Code as the entrypoint) - README.md: rewrite to reflect the new entrypoint and upstream URL Signed-off-by: Eli Aleyner <eli@aleyner.com>
Replace the local dev path (./nanoclaw12/) with the canonical git+https kit URL so users can copy-paste the example directly. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Extend the install step to also build the nanoclaw agent container image (./container/build.sh) so the kit is fully self-contained on first boot. Add `set -e` so partial installs don't unblock the entrypoint by writing .installed prematurely. Harden the entrypoint: - chmod the docker socket on each boot (sbx restarts dockerd, which resets permissions) - tail the install log to stderr while waiting on first boot so the user sees progress during the 5-12 minute build - redirect service stdout/stderr to logs/nanoclaw.log and logs/nanoclaw.error.log instead of a single combined log - wait up to 60s for data/cli.sock and dump both service logs on timeout, so failures are diagnosable without exec'ing in Add egress hosts needed by the container build (api.onecli.sh, deb.debian.org, security.debian.org) and bind onecli to 0.0.0.0 with NO_PROXY set so the agent container can reach it via gateway.docker.internal. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Signed-off-by: Eli Aleyner <eli@aleyner.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
qwibitai/nanoclawtonanocoai/nanoclaw.~/.local/binonPATH.spec.yaml
github.com/nanocoai/nanoclaw; install viapnpm install --network-concurrency=1.~/.local/bininPATHvia/etc/sandbox-persistent.shso OneCLI's post-install check works in every shell context.nodejs.org,onecli.sh,www.onecli.sh,raw.githubusercontent.com,release-assets.githubusercontent.com,ghcr.io,registry-1.docker.io,auth.docker.io,production.cloudflare.docker.com,checkpoint.prisma.io.NODE_NO_WARNINGS=1andONECLI_BIND_HOST=127.0.0.1.data/cli.sock, then runsnpm run setup:autowithNANOCLAW_SKIP=service,container.aiFilename: CLAUDE.mdandsettings.containerSettings.claude— the kit no longer uses Claude Code as the entrypoint.README.md
/add-<channel>skills from inside the session.Test plan
sbx kit validate ./nanoclaw/— passesgo test -v -count=1 -timeout 10m ./...innanoclaw/— all validation/policy/oauth subtests pass; thecontainersubtest requires rootless Docker locally and is left to CIsbx run --kit ./nanoclaw/ nanoclawend-to-end (verify install completes, service starts, setup wizard appears on attach)