coldpass

JHNMACHINE/moonclip

GitHub ↗ · How this check works

Per-setup public labels: Linux / Haiku: Works

Static status badge · SVG · read-only JSON. Embed with ![Coldpass status](https://coldpass.dev/badge/JHNMACHINE--moonclip.svg).

Works

Only older checks are saved; no current product test.

Answers apply to each recorded setup, revision and operation. Earlier observations remain in Run details.

Tested on Linux
PlatformHaikuLuna
Linuxyes

Linux / Haiku 5.5: yes · Oct 18, 2026 · a856d0ddf1c4 · claude-code / 2.1.300 (Claude Code)
Recorded operation: <host-path> bench/benchmark_checkpoints.py --n-saves 2 --methods moonclip --output-dir <host-path>

Needs keys or browser
None seen
Issue
none found in recorded check

Safety: no observations were recorded for the selected revision. Functional results do not establish a safety result.

Run details

Haiku 5.5

Native agent · Checked operation passed

Checked operation <host-path> bench/benchmark_checkpoints.py --n-saves 2 --methods moonclip --output-dir <host-path>

Agent-reported scope: Ran bench/benchmark_checkpoints.py (README 'Benchmarks' section) for moonclip: 2 saves of MiniGPT 41.7M params, reloaded and verified (OK check passed).

Run details
run
JHNMACHINE--moonclip-claude-haiku-5-5-1792340931569239804
date
Oct 18, 2026
time to working
not recorded
product hosting / execution
Local command in the guest
recorded operation scope
product operation
model
claude-haiku-5-5
agent / version
claude-code / 2.1.300 (Claude Code)
platform
linux
reported cost
$0.007820
Recorded source
https://github.com/JHNMACHINE/moonclip
Tool transport
mcp-assigned-vm
Test version
readme-03aea4c41d53955e3723c330e9927d2a9df627f6642624cc0078a97c73a53c8a
revision
a856d0ddf1c4eb544b04fdf285383c0e5f99ea05
What the agent did

Ran bench/benchmark_checkpoints.py (README 'Benchmarks' section) for moonclip: 2 saves of MiniGPT 41.7M params, reloaded and verified (OK check passed).

cd <host-path> && ls -la && cat README* 2>/dev/null | head -300
cd <host-path> && git log -1 --oneline 2>/dev/null; ls .github/workflows 2>/dev/null; echo "CDP=$CDP_URL $<env-var>"; which node npm python3 pip3 uv 2>&1
uname -m; python3 --version; python3 -m pip --version 2>&1 | head -2; cd /tmp && python3 -m venv <host-path> 2>&1 | tail -3; <host-path> install -q --upgrade pip 2>&1 | tail -3; <host-path> install moonclip 2>&1 | tail -15
sudo DEBIAN_FRONTEND=noninteractive apt-get install -y -q python3-venv python3-pip >/tmp/apt.log 2>&1; echo "apt rc=$?"; tail -3 /tmp/apt.log; rm -rf <host-path>; python3 -m venv <host-path> && <host-path> install -q --upgrade pip 2>&1 | tail -3; <host-path> install moonclip 2>&1 | tail -15
sudo apt-get update -q >/tmp/aptu.log 2>&1; echo "update rc=$?"; sudo DEBIAN_FRONTEND=noninteractive apt-get install -y -q python3.12-venv python3-pip >/tmp/apt.log 2>&1; echo "apt rc=$?"; tail -3 /tmp/apt.log; rm -rf <host-path>; python3 -m venv <host-path> && <host-path> install -q --upgrade pip 2>&1 | tail -3; <host-path> install moonclip 2>&1 | tail -15
cd <host-path> && ls examples bench; <host-path> -c "import torch" 2>&1 | tail -1; grep -n "run:" -A3 .github/workflows/checks.yml | head -40; <host-path> -c "import moonclip,sys;print(moonclip.__file__, dir(moonclip))"
cd <host-path> && export NODE_USE_ENV_PROXY=1 && timeout 900 <host-path> install -q torch --index-url https://download.pytorch.org/whl/cpu > /tmp/torch.log 2>&1; echo "rc=$?"; tail -3 /tmp/torch.log; <host-path> -c "import torch;print(torch.__version__)" 2>&1 | tail -1; head -40 bench/benchmark_checkpoints.py
cd <host-path> && timeout 900 <host-path> install -q safetensors accelerate pytorch-lightning > /tmp/extra.log 2>&1; echo "rc=$?"; tail -3 /tmp/extra.log; grep -n "add_argument\|^import\|^from" bench/benchmark_checkpoints.py | head -30; grep -n "out\|dir" bench/benchmark_checkpoints.py | grep -i "default\|add_argument\|tmp\|mkdtemp" | head
cd <host-path> && sed -n 705,760p bench/benchmark_checkpoints.py; ls -d bench_checkpoints 2>/dev/null; echo "---"; timeout 900 <host-path> bench/benchmark_checkpoints.py --n-saves 2 --methods moonclip --output-dir <host-path> > /tmp/bench.log 2>&1; echo "rc=$?"; tail -30 /tmp/bench.log
cd <host-path> && mkdir -p .coldpass && rm -rf .coldpass/bench_out && cat > .coldpass/probe.json <<'EOF'
{"kind":"cli","argv":["<host-path>","stdin":"","env":{}}
EOF
cat > .coldpass/report.json <<'EOF'
{"outcome":"working","operation":"Ran bench/benchmark_checkpoints.py (README 'Benchmarks' section) for moonclip: 2 saves of MiniGPT 41.7M params, reloaded and verified (OK check passed).","operation_scope":"Moonclip checkpoint save/load on CPU only, moonclip method only (no torch.save/safetensors comparison runs). Installed from PyPI wheel, not built from the pinned commit.","steps":["apt-get install python3.12-venv python3-pip (after apt-get update)","python3 -m venv <host-path>","pip install --upgrade pip","pip install moonclip (0.2.3 wheel, README Installation)","pip install torch --index-url https://download.pytorch.org/whl/cpu (2.14.1+cpu)","pip install safetensors accelerate pytorch-lightning (benchmark usage line)","python bench/benchmark_checkpoints.py --n-saves 2 --methods moonclip --output-dir .../.coldpass/bench_out"],"deviations":["README 'Installation' says pip install moonclip; the pinned commit a856d0d was not built from source (needs maturin and a Rust 1.98.1 toolchain). The PyPI wheel 0.2.3 was used instead.","Guest had no ensurepip: python3 -m venv failed; installed python3.12-venv and python3-pip via apt.","torch was installed from the CPU index, since the benchmark needs torch and the guest has no GPU.","Did not run cargo test, clippy, pytest or other CI suites, per task instructions."],"stuck_step":null,"main_blocker":null,"suggested_fix":null}
EOF
<env-var>=<host-path> python3 -I <host-path> --validate; echo "rc=$?"