Back to skill

Security audit

Snapmaker U1

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill benign with medium confidence: the checked evidence supports its stated 3D modeling and print preparation purpose, while noting an unpinned installer command and incomplete coverage.

The verdict is benign, but it is not a guarantee of safety. Consider whether you are comfortable installing through an unpinned `npx` package, which may change between installs. Coverage is marked partial: one local reference was unresolved and binary content was outside some analyzers' scope, so the evidence is incomplete.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
Findings (37)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The supplied chunk contains only a skill-installation command. It implements none of the described geometry, toolpath, color, print-settings, or printer-query functionality. Although installation could precede use of a matching skill, the supplied code's sole behavior is external skill installation; the functionality of the installed skill cannot be established from this chunk.

Content

No source excerpt is available for this finding.

Self-Modification

High
Category
Rogue Agent
Confidence
85% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · scripts/u1_vision.swift (reported line 70)May include surrounding context.

text
for path in [output, report].compactMap({ $0 }) {
            // attributesOfItem also identifies existing symbolic links.
            if (try? FileManager.default.attributesOfItem(atPath: path.path)) != nil {
                throw VisionCLIError(code: "output_exists", message: "Refusing to overwrite existing file: \(path.path)")
            }
        }
    }

MCP server referenced without pinned version: 'npx skills'.

Medium
Category
MCP Rug Pull
Confidence
94% confidence
Finding

The installation command invokes an unpinned npm package through npx, allowing its resolved release to change and execute with the installing user's privileges. This is a supply-chain exposure rather than an MCP-server issue; compromise of a future package release could affect local files and credentials even though the skill's intended workflow is local modeling.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, file_read, file_write, network, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 38)May include surrounding context.

md
## Geometry and print plan

Preserve originals. Inspect and repair only unintended defects; record changes and recheck form. Confirm units, axis, dimensions, stable bed contact, thin features, disconnected shells, winding and self-intersections where supported. Do not remove all but the largest component without checking whether details or moving parts are intentional.

Run `scripts/u1_mesh_check.py` on every delivered STL. `scripts/u1_prepare.py` also screens the serialized STL/GLB bytes: float32 export can collapse triangles that passed an in-memory check. The bundled checker cannot certify likeness, self-intersections, strength or physical fit.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_bambu.py (reported line 60)May include surrounding context.

python
def verify_runtime(runtime):
    """Use only the reviewed revision with unchanged code, references and rules."""
    try:
        revision = subprocess.check_output(["git", "-C", str(runtime), "rev-parse", "HEAD"], text=True, stderr=subprocess.STDOUT).strip()
        changes = subprocess.check_output(["git", "-C", str(runtime), "status", "--porcelain", "--untracked-files=all", "--", "backend", "profiles", "rules", "LICENSE"], text=True, stderr=subprocess.STDOUT).strip()
    except (OSError, subprocess.CalledProcessError) as exc:
        raise ConversionError("Local Bambu converter is missing; see references/BAMBU_CONVERSION.md for setup.") from exc

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_bambu.py (reported line 61)May include surrounding context.

python
"""Use only the reviewed revision with unchanged code, references and rules."""
    try:
        revision = subprocess.check_output(["git", "-C", str(runtime), "rev-parse", "HEAD"], text=True, stderr=subprocess.STDOUT).strip()
        changes = subprocess.check_output(["git", "-C", str(runtime), "status", "--porcelain", "--untracked-files=all", "--", "backend", "profiles", "rules", "LICENSE"], text=True, stderr=subprocess.STDOUT).strip()
    except (OSError, subprocess.CalledProcessError) as exc:
        raise ConversionError("Local Bambu converter is missing; see references/BAMBU_CONVERSION.md for setup.") from exc
    if revision != COMMIT or changes:

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_bambu.py (reported line 487)May include surrounding context.

python
command = [str(python), "-I", str(ROOT / "scripts/u1_offline_exec.py"), "--", str(Path(__file__).absolute()), "_worker", operation, str(runtime)]
    if request_path:
        command.append(str(request_path))
    completed = subprocess.run(command, text=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE, timeout=600, check=False)
    if completed.returncode:
        raise ConversionError(completed.stderr.strip()[-4000:] or "Local conversion worker failed.")
    return json.loads(completed.stdout)

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_local.py (reported line 103)May include surrounding context.

python
truststore.inject_into_ssl()
    from huggingface_hub import HfApi, hf_hub_download
    if not source.exists():
        subprocess.run(['git', 'init', str(source)], check=True, stdout=subprocess.DEVNULL)
        subprocess.run(['git', '-C', str(source), 'remote', 'add', 'origin', SOURCE_URL], check=True)
        subprocess.run(['git', '-C', str(source), 'fetch', '--depth', '1', 'origin', SOURCE_COMMIT], check=True)
        subprocess.run(['git', '-C', str(source), 'checkout', '--detach', 'FETCH_HEAD'], check=True)

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_local.py (reported line 104)May include surrounding context.

python
from huggingface_hub import HfApi, hf_hub_download
    if not source.exists():
        subprocess.run(['git', 'init', str(source)], check=True, stdout=subprocess.DEVNULL)
        subprocess.run(['git', '-C', str(source), 'remote', 'add', 'origin', SOURCE_URL], check=True)
        subprocess.run(['git', '-C', str(source), 'fetch', '--depth', '1', 'origin', SOURCE_COMMIT], check=True)
        subprocess.run(['git', '-C', str(source), 'checkout', '--detach', 'FETCH_HEAD'], check=True)
    current = subprocess.check_output(['git', '-C', str(source), 'rev-parse', 'HEAD'], text=True).strip()

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_local.py (reported line 105)May include surrounding context.

python
if not source.exists():
        subprocess.run(['git', 'init', str(source)], check=True, stdout=subprocess.DEVNULL)
        subprocess.run(['git', '-C', str(source), 'remote', 'add', 'origin', SOURCE_URL], check=True)
        subprocess.run(['git', '-C', str(source), 'fetch', '--depth', '1', 'origin', SOURCE_COMMIT], check=True)
        subprocess.run(['git', '-C', str(source), 'checkout', '--detach', 'FETCH_HEAD'], check=True)
    current = subprocess.check_output(['git', '-C', str(source), 'rev-parse', 'HEAD'], text=True).strip()
    if current != SOURCE_COMMIT:

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_local.py (reported line 106)May include surrounding context.

python
subprocess.run(['git', 'init', str(source)], check=True, stdout=subprocess.DEVNULL)
        subprocess.run(['git', '-C', str(source), 'remote', 'add', 'origin', SOURCE_URL], check=True)
        subprocess.run(['git', '-C', str(source), 'fetch', '--depth', '1', 'origin', SOURCE_COMMIT], check=True)
        subprocess.run(['git', '-C', str(source), 'checkout', '--detach', 'FETCH_HEAD'], check=True)
    current = subprocess.check_output(['git', '-C', str(source), 'rev-parse', 'HEAD'], text=True).strip()
    if current != SOURCE_COMMIT:
        raise ValueError('Runtime checkout differs from the reviewed revision; choose a fresh runtime directory.')

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_local.py (reported line 107)May include surrounding context.

python
subprocess.run(['git', '-C', str(source), 'remote', 'add', 'origin', SOURCE_URL], check=True)
        subprocess.run(['git', '-C', str(source), 'fetch', '--depth', '1', 'origin', SOURCE_COMMIT], check=True)
        subprocess.run(['git', '-C', str(source), 'checkout', '--detach', 'FETCH_HEAD'], check=True)
    current = subprocess.check_output(['git', '-C', str(source), 'rev-parse', 'HEAD'], text=True).strip()
    if current != SOURCE_COMMIT:
        raise ValueError('Runtime checkout differs from the reviewed revision; choose a fresh runtime directory.')
    iso = source / 'tsr/models/isosurface.py'

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_local.py (reported line 111)May include surrounding context.

python
if current != SOURCE_COMMIT:
        raise ValueError('Runtime checkout differs from the reviewed revision; choose a fresh runtime directory.')
    iso = source / 'tsr/models/isosurface.py'
    pristine = subprocess.check_output(['git', '-C', str(source), 'show', 'HEAD:tsr/models/isosurface.py'], text=True)
    expected = pristine.replace(PATCH_FROM, PATCH_TO)
    if pristine == expected or iso.read_text() not in [pristine, expected]:
        raise ValueError('Unexpected isosurface source; refusing to overwrite changes.')

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_local.py (reported line 138)May include surrounding context.

python
# Upstream tokenizer calls hf_hub_download. A narrow local-only patch uses its
    # existing config path, not network downloads during image processing.
    tokenizer = source / 'tsr/models/tokenizers/image.py'
    raw = subprocess.check_output(['git', '-C', str(source), 'show', 'HEAD:tsr/models/tokenizers/image.py'], text=True)
    old_call = '''hf_hub_download(
                    repo_id=self.cfg.pretrained_model_name_or_path,
                    filename="config.json",

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_local.py (reported line 229)May include surrounding context.

python
raise ValueError('Apple Vision helper is not built; compile scripts/u1_vision.swift first.')
            native_output = output / 'vision_subject.png'
            native_json = output / 'vision_report.json'
            completed = subprocess.run([str(native_binary), '--input', str(Path(image_path).resolve()),
                                        '--output', str(native_output.resolve()), '--report', str(native_json.resolve())],
                                       capture_output=True, text=True, timeout=120, check=False)
            if completed.returncode != 0:

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · scripts/u1_slice.py (reported line 58)May include surrounding context.

python
def resolve_preset(profiles_dir: str | Path, kind: str, name: str, *, _chain: tuple[str, ...] = ()) -> tuple[dict, list[dict]]:
    """Resolve one vendor's inheritance without asking the CLI to find parents."""
    if kind not in {"machine", "process", "filament"}:
        raise SliceError(f"unsupported preset type: {kind}")
    if not name or name in {".", ".."} or any(char in name for char in "/\\"):

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_slice.py (reported line 299)May include surrounding context.

python
else:
        with transcript.open("w", encoding="utf-8") as stream:
            try:
                completed = subprocess.run(argv, cwd=run_dir, stdout=stream, stderr=subprocess.STDOUT, timeout=timeout, check=False)
                return_code = completed.returncode
            except subprocess.TimeoutExpired:
                timed_out = True

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_trellis.py (reported line 65)May include surrounding context.

python
source = runtime / 'trellis-v1'
    dino = runtime / 'torch/hub/facebookresearch_dinov2_main'
    for checkout, expected in [(source, SOURCE_COMMIT), (dino, DINO_COMMIT)]:
        actual = subprocess.check_output(['git', '-C', str(checkout), 'rev-parse', 'HEAD'], text=True).strip()
        if actual != expected:
            raise ValueError('Unexpected source revision: ' + str(checkout))
    public = json.loads((source / 'public_weights.json').read_text())

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_trellis.py (reported line 189)May include surrounding context.

python
if not vision.is_file() or not os.access(vision, os.X_OK):
                raise ValueError('Opaque input requires the configured native Apple Vision executable.')
            report = folder / f'vision-{index:02d}.json'
            result = subprocess.run([str(vision), '--input', str(saved), '--output', str(target),
                                     '--report', str(report), '--mode', 'foreground'], capture_output=True, text=True, timeout=120)
            if result.returncode:
                raise ValueError('Native extraction failed: ' + (result.stdout or result.stderr).strip()[:800])

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/u1_trellis.py (reported line 243)May include surrounding context.

python
write_json(output / 'job.json', job)
        print('Local progress log: ' + str(output / 'run.log'), flush=True)
        with (output / 'run.log').open('w') as log:
            completed = subprocess.run(command, cwd=manifest['source'], env=offline_environment(runtime, manifest),
                                       stdout=log, stderr=subprocess.STDOUT)
        if completed.returncode or not (output / 'model.glb').is_file():
            raise RuntimeError(f'Local generation failed (exit {completed.returncode}); inspect run.log. Earlier raw/latent outputs are preserved.')

Tainted flow: 'command' from pathlib.Path.read_text (line 237, file read) → subprocess.run (code execution)

Medium
Category
Data Flow
Confidence
65% confidence
Finding

Data from a source is assigned to a variable that is later passed to a sink, creating a variable-mediated taint flow.

Content

Scanner excerpt · scripts/u1_trellis.py (reported line 243)May include surrounding context.

python
write_json(output / 'job.json', job)
        print('Local progress log: ' + str(output / 'run.log'), flush=True)
        with (output / 'run.log').open('w') as log:
            completed = subprocess.run(command, cwd=manifest['source'], env=offline_environment(runtime, manifest),
                                       stdout=log, stderr=subprocess.STDOUT)
        if completed.returncode or not (output / 'model.glb').is_file():
            raise RuntimeError(f'Local generation failed (exit {completed.returncode}); inspect run.log. Earlier raw/latent outputs are preserved.')

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · tests/test_u1_apple_plan.py (reported line 32)May include surrounding context.

python
cls.binary = Path(cls.temporary.name) / "u1_apple_plan"
        command = ["xcrun", "swiftc", "-parse-as-library", "-module-cache-path",
                   str(Path(cls.temporary.name) / "module-cache"), str(SOURCE), "-o", str(cls.binary)]
        built = subprocess.run(command, capture_output=True, text=True, timeout=120)
        if built.returncode:
            unsupported = ("sandbox_apply", "no such module 'FoundationModels'", "SDK is not supported by the compiler",
                           "unable to find utility", "unable to load standard library")

Tainted flow: 'command' from os.environ.get (line 30, credential/environment) → subprocess.run (code execution)

Medium
Category
Data Flow
Confidence
65% confidence
Finding

Data from a source is assigned to a variable that is later passed to a sink, creating a variable-mediated taint flow.

Content

Scanner excerpt · tests/test_u1_apple_plan.py (reported line 32)May include surrounding context.

python
cls.binary = Path(cls.temporary.name) / "u1_apple_plan"
        command = ["xcrun", "swiftc", "-parse-as-library", "-module-cache-path",
                   str(Path(cls.temporary.name) / "module-cache"), str(SOURCE), "-o", str(cls.binary)]
        built = subprocess.run(command, capture_output=True, text=True, timeout=120)
        if built.returncode:
            unsupported = ("sandbox_apply", "no such module 'FoundationModels'", "SDK is not supported by the compiler",
                           "unable to find utility", "unable to load standard library")

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · tests/test_u1_apple_plan.py (reported line 41)May include surrounding context.

python
raise AssertionError("Swift helper failed to compile:\n" + built.stderr)

    def invoke(self, *arguments):
        completed = subprocess.run([str(self.binary), *arguments], capture_output=True, text=True, timeout=30)
        self.assertTrue(completed.stdout.strip(), completed.stderr)
        try:
            report = json.loads(completed.stdout)