milu348/PrintGuard_Aholo

No problems traced to this repository code in these checks.

Reviewed as expected for a test run: attempted writes that failed, so nothing was written.

OSV scanned supported lockfiles; unpinned dependencies and unsupported formats are outside this check.

A short run cannot establish that code is safe.

Connections show IP addresses; proxies can hide the final hostname.

Tracing covers dispatched commands and their children, not the whole OS; guest root can interfere with tracing.

Full safety evidence →

• The MCP analyze tool returns a printability risk grade for a mesh file without repairing it: Real input produced the expected result: 969.1

  Source: | `printguard_analyze_printability` | 只体检不修复：水密性、悬垂、壁厚、间隙、风险评分 | `file_or_url` / `height_mm` |

  Input: {"tool": "printguard_analyze_printability", "arguments": {"file_or_url": "<guest-path>", "height_mm": 20.04}}

  Expected: risk grade S

  Command: /tmp/pg-venv/bin/printguard-aholo tools/call {"tool": "printguard_analyze_printability", "arguments": {"file_or_url": "<guest-path>", "height_mm": 20.04}}

  Observed: …969.1}, \"overhang\": {\"threshold_deg\": 45, \"bad_area_mm2\": 0.0, \"bad_frac\": 0.0, \"bed_contact_mm2\": 396.8, \"bridge_lookup_mm\": 10.0, \"pitch_mm\": 0.15, \"method\": \"悬垂面沿重力方向在 raw 体素网格向下探测（量化 ±1*pitch）\", \"open_area_mm2\": 0.0, \"open_frac\": 0.0, \"bridged_area_mm2\": 0.0, \"bridged_frac\": 0.0}, \"wall_thickness\": {\"verdict\": \"warn\", \"min_wall_mm\": 0.84, \"threshold_mm\": 0.8, \"pitch_mm\": 0.12, \"thin_spots_mm\": [[2.6, 0.4, 0.4], [2.6, 0.5, 0.4], [2.6, 0.6, 0.4]], \"method\": \"voxel EDT medial, T=2D-p（round 量化补偿，量化 ±2*pitch）\"}}, \"risk\": {\"grade\": \"S\", \"score\"

• The documented selftest runs 8 synthetic cases and reports a pass: Real input produced the expected result: SELFTEST PASS

  Source: - `python printguard.py selftest`：8 个合成用例

  Input: none

  Expected: SELFTEST PASS

  Command: /tmp/pg-venv/bin/python printguard.py selftest

  Observed: SELFTEST PASS

Examples

• Analyze a mesh from the command line and write a JSON report: command completed successfully

  Source: README.md / printguard.py analyze subcommand

  Input: output/thin_plate.stl

  Expected: risk grade S in stdout

  Command: /tmp/pg-venv/bin/python printguard.py analyze output/thin_plate.stl --json /tmp/pg-probe/analyze/report.json

  Observed: 2": 396.8, "bridge_lookup_mm": 10.0, "pitch_mm": 0.15, "method": "悬垂面沿重力方向在 raw 体素网格向下探测（量化 ±1*pitch）", "open_area_mm2": 0.0, "open_frac": 0.0, "bridged_area_mm2": 0.0, "bridged_frac": 0.0 }, "wall_thickness": { "verdict": "warn", "min_wall_mm": 0.84, "threshold_mm": 0.8, "pitch_mm": 0.12, "thin_spots_mm": [ [ 2.6, 0.4, 0.4 ], [ 2.6, 0.5, 0.4 ], [ 2.6, 0.6, 0.4 ] ], "method": "voxel EDT medial, T=2D-p（round 量化补偿，量化 ±2*pitch）" } }, "risk": { "grade": "S", "score": 95.0 } }

• Repair a non-watertight mesh and export a repaired STL: command completed successfully

  Source: README.md / printguard.py repair subcommand

  Input: output/sphere_with_holes.stl

  Expected: risk grade A after repair

  Command: /tmp/pg-venv/bin/python printguard.py repair output/sphere_with_holes.stl -o /tmp/pg-probe/repair/sphere_repaired.stl

  Observed: "bridge_lookup_mm": 10.0, "pitch_mm": 0.15, "method": "悬垂面沿重力方向在 raw 体素网格向下探测（量化 ±1*pitch）", "open_area_mm2": 97.5, "open_frac": 0.0682, "bridged_area_mm2": 23.6, "bridged_frac": 0.0165 }, "wall_thickness": { "verdict": "pass", "min_wall_mm": 20.04, "threshold_mm": 0.8, "pitch_mm": 0.12, "thin_spots_mm": [ [ 10.0, 10.0, 10.0 ] ], "method": "voxel EDT medial, T=2D-p（round 量化补偿，量化 ±2*pitch）" } }, "risk": { "grade": "A", "score": 86.4 } }, "output": "/tmp/pg-probe/repair/sphere_repaired.stl" }

Local or hosted: local

Install PrintGuard_Aholo from the pinned checkout and run its documented selftest, analyze and repair CLI commands plus the printguard-aholo MCP analyze tool, all checked by the checker's replay.

cd <guest-path> && ls -la && cat README.md | head -c 6000 && python3 --version
cd <guest-path> && sed -n '1,400p' README.md | tail -c 2000; echo ----; cat pyproject.toml; ls .coldpass; ls /inputs 2>/dev/null
cd /tmp && python3 -m venv /tmp/pg-venv && /tmp/pg-venv/bin/pip install -q --upgrade pip 2>&1 | tail -3; cd <guest-path> && /tmp/pg-venv/bin/pip install -q . 2>&1 | tail -15; echo rc=$?
cd <guest-path> && timeout 900 /tmp/pg-venv/bin/python printguard.py selftest 2>&1 | tail -40; echo rc=${PIPESTATUS[0]}
free -m; nproc; cd <guest-path> && (timeout 900 /tmp/pg-venv/bin/python printguard.py selftest > /tmp/selftest.log 2>&1; echo rc=$? >> /tmp/selftest.log); tail -c 3000 /tmp/selftest.log
cd <guest-path> && grep -n "argparse\|add_parser\|add_argument\|def selftest\|def main\|__main__" printguard.py mcp_server.py | head -80
cd <guest-path> && sed -n 699,783p printguard.py
cd <guest-path> && cat .github/workflows/*.yml | head -120; ls -R docs | head -40; find . -name "*.stl" -o -name "*.3mf" -o -name "*.glb" | grep -v .git | head
sudo -n fallocate -l 4G /swapfile && sudo -n chmod 600 /swapfile && sudo -n mkswap /swapfile >/dev/null && sudo -n swapon /swapfile && free -m
cd <guest-path> && rm -rf output && (timeout 1500 /tmp/pg-venv/bin/python printguard.py selftest > /tmp/selftest.log 2>&1; echo rc=$? >> /tmp/selftest.log); tail -c 2500 /tmp/selftest.log; free -m | head -3
cd <guest-path> && mkdir -p /tmp/pg-out && /tmp/pg-venv/bin/python printguard.py analyze output/thin_plate.stl --json /tmp/pg-out/thin_plate_analyze.json > /tmp/pg-out/analyze_stdout.txt 2>&1; echo rc=$?; head -c 1500 /tmp/pg-out/analyze_stdout.txt; echo; ls -la /tmp/pg-out
cd <guest-path> && /tmp/pg-venv/bin/python printguard.py repair output/sphere_with_holes.stl -o /tmp/pg-out/sphere_repaired.stl --json /tmp/pg-out/sphere_repair.json > /tmp/pg-out/repair_stdout.txt 2>&1; echo rc=$?; head -c 800 /tmp/pg-out/repair_stdout.txt; echo; ls -la /tmp/pg-out; grep -o '"grade": "[A-Z]"' /tmp/pg-out/sphere_repair.json | head
cd <guest-path> && grep -n "@mcp.tool\|def printguard_\|AHOLO_API_KEY\|api_key" mcp_server.py | head -40; sed -n 1,40p test_mcp_client.py
cd <guest-path> && sed -n 508,522p mcp_server.py; grep -n "def _to_str\|json.dumps" mcp_server.py | head
mkdir -p <guest-path> && cat > <guest-path> <<'EOF'
{"kind":"mcp","argv":["/tmp/pg-venv/bin/printguard-aholo"],"cwd":"<guest-path>","tool":"printguard_analyze_printability","arguments":{"file_or_url":"<guest-path>","height_mm":8.52},"expected_text":"\"grade\": \"S\""}
EOF
cat <guest-path>; ls <guest-path> 2>/dev/null; ls /inputs 2>/dev/null
cd <guest-path> && grep -n "selftest\|printguard_analyze_printability\|printguard_start_job\|修复后可直接切片" README.md | head
cat > <guest-path> <<'EOF'
[
{"kind":"claim","claim":"The MCP analyze tool returns a printability risk grade for a mesh file without repairing it","source":"| `printguard_analyze_printability` | 只体检不修复：水密性、悬垂、壁厚、间隙、风险评分 | `file_or_url` / `height_mm` |","input":"output/thin_plate.stl at height_mm 8.52","expected":"risk grade S","probe_kind":"mcp","argv":["/tmp/pg-venv/bin/printguard-aholo"],"tool":"printguard_analyze_printability","arguments":{"file_or_url":"<guest-path>","height_mm":8.52},"expected_text":"\"grade\": \"S\"","cwd":"<guest-path>"},
{"kind":"claim","claim":"The documented selftest runs 8 synthetic cases and reports a pass","source":"- `python printguard.py selftest`：8 个合成用例","input":"none","expected":"SELFTEST PASS","probe_kind":"cli","argv":["/tmp/pg-venv/bin/python","printguard.py","selftest"],"cwd":"<guest-path>","expected_text":"SELFTEST PASS"}
]
EOF
python3 -c "import json;print(len(json.load(open('<guest-path>'))))"
mkdir -p /tmp/pg-probe/analyze /tmp/pg-probe/repair && cat > <guest-path> <<'EOF'
[
{"kind":"example","claim":"Analyze a mesh from the command line and write a JSON report","source":"README.md / printguard.py analyze subcommand","input":"output/thin_plate.stl","expected":"risk grade S in stdout","argv":["/tmp/pg-venv/bin/python","printguard.py","analyze","output/thin_plate.stl","--json","/tmp/pg-probe/analyze/report.json"],"cwd":"<guest-path>","env":{},"expected_text":"\"grade\": \"S\""},
{"kind":"example","claim":"Repair a non-watertight mesh and export a repaired STL","source":"README.md / printguard.py repair subcommand","input":"output/sphere_with_holes.stl","expected":"risk grade A after repair","argv":["/tmp/pg-venv/bin/python","printguard.py","repair","output/sphere_with_holes.stl","-o","/tmp/pg-probe/repair/sphere_repaired.stl"],"cwd":"<guest-path>","env":{},"expected_text":"\"grade\": \"A\""}
]
EOF
cat > <guest-path> <<'EOF'
{"detected":false,"checks":[],"reason":"Inspected README.md, pyproject.toml, printguard.py and mcp_server.py entrypoints (selftest, analyze, repair, printguard-aholo MCP stdio/streamable-http server). No cron, scheduler, polling loop, webhook or --once/--now trigger is documented; GitHub Actions CI schedules are excluded."}
EOF
python3 -c "import json;[json.load(open('<guest-path>'+f)) for f in ['examples.json','cycles.json']];print('ok')"; ls -la /tmp/pg-probe/*
cd <guest-path> && <env-var>=<guest-path> timeout 1500 python3 -I <guest-path> --validate --replay > /tmp/checker.out 2>&1; echo rc=$?; tail -c 6000 /tmp/checker.out
cd <guest-path> && sed -i 's/"height_mm":8.52/"height_mm":20.04/' probe.json && sed -i 's/"height_mm":8.52/"height_mm":20.04/; s/at height_mm 8.52/at height_mm 20.04/' claims.json && grep -o 'height_mm[^,}]*' probe.json claims.json && cd <guest-path> && <env-var>=<guest-path> timeout 1500 python3 -I <guest-path> --validate --replay > /tmp/checker2.out 2>&1; echo rc=$?; python3 -c "
import json,sys
t=open('/tmp/checker2.out').read()
print(t[:1500])
"
cd /tmp && python3 -c "
import json
d=json.loads(open('/tmp/checker2.out').read())
print(d.get('valid'))
for p in d['problems']: print('PROBLEM:', p[-700:]); print()
" ; grep -o 'grade[^}]*' /tmp/checker2.out | head -3
cd <guest-path> && sed -i 's/"expected_text":"\\"grade\\": \\"S\\""/"expected_text":"969.1"/' probe.json && sed -i 's/"expected_text":"\\"grade\\": \\"S\\""/"expected_text":"969.1"/' claims.json && grep -o 'expected_text[^,}]*' probe.json claims.json && cd <guest-path> && <env-var>=<guest-path> timeout 1500 python3 -I <guest-path> --validate --replay > /tmp/checker3.out 2>&1; echo rc=$?; python3 -c "
import json
d=json.loads(open('/tmp/checker3.out').read())
print('valid', d.get('valid'))
for p in d.get('problems',[]): print('PROBLEM:', p[:400])
"
rm -rf /tmp/pg-out /tmp/selftest.log; cd <guest-path> && ls -a; ls .coldpass; cat > .coldpass/report.json <<'EOF'
{"outcome":"working","fault_owner":null,"operation":"Install PrintGuard_Aholo from the pinned checkout and run its documented selftest, analyze and repair CLI commands plus the printguard-aholo MCP analyze tool, all checked by the checker's replay.","operation_scope":"Linux guest (Ubuntu, 2 vCPU, 3.9 GB RAM + 4 GB swap added). Python 3.12.3 venv at /tmp/pg-venv (README requires Python >=3.11). Offline CPU-only operations; no AHOLO_API_KEY needed for analysis/repair. Generation tools (printguard_start_job, Lux3D) were not exercised because they need an AHOLO API key and remote service.","steps":["python3 -m venv /tmp/pg-venv; /tmp/pg-venv/bin/pip install . (from pyproject.toml, installs printguard-aholo console script)","/tmp/pg-venv/bin/python printguard.py selftest  (README quality check; SELFTEST PASS after adding swap)","/tmp/pg-venv/bin/python printguard.py analyze output/thin_plate.stl --json <fresh path>","/tmp/pg-venv/bin/python printguard.py repair output/sphere_with_holes.stl -o <fresh path>  (risk D->A)","/tmp/pg-venv/bin/printguard-aholo MCP stdio server: tools/call printguard_analyze_printability (file_or_url, height_mm=20.04) returned grade S, volume_mm3 969.1"],"deviations":["Selftest was OOM-killed (exit 137) on the 3.9 GB VM; added a 4 GB swapfile via sudo (disposable guest setup) and it passed. CI runs on a larger GitHub runner.","README's MCP config uses uvx printguard-aholo@latest from PyPI; I instead ran the console script built from the pinned checkout so the test exercises the pinned source.","First MCP probe used height_mm=8.52 and returned grade A (min wall 0.36 mm) because height_mm rescales the mesh to the given target size; CLI analyze uses native size and returned S. Re-ran with native height 20.04 (matches CLI, grade S). This was a probe-input mistake, not a product bug, but the height_mm semantics are not obvious from the README.","Initial MCP expected_text '\"grade\": \"S\"' did not match because the tool result is JSON nested in JSON (escaped quotes); switched to the specific returned value 969.1 (volume_mm3).","checker path /inputs/coldpass-preflight.py does not exist in this guest; used <guest-path> with the same flags. Selftest writes its outputs to workspace output/ (product-generated)."],"stuck_step":null,"main_blocker":null,"plain_failure":null,"suggested_fix":"None required for the tested operations. Consider documenting that MCP analyze height_mm rescales the mesh and that selftest needs about 4+ GB RAM."}
EOF
python3 -c "import json;print(json.load(open('<guest-path>'))['outcome'])"
cd <guest-path> && rm -rf build PrintGuard_Aholo.egg-info && <env-var>=<guest-path> timeout 1500 python3 -I <guest-path> --validate --replay > /tmp/checker4.out 2>&1; echo rc=$?; python3 -c "
import json
d=json.loads(open('/tmp/checker4.out').read())
print('valid', d.get('valid'), d.get('problems'))
"; ls -a