Back to skill

Security audit

Package Detection Skill | 包裹检测技能

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marked this skill suspicious because scanners report credential and identity handling, remote requests, and behaviors beyond its stated package-detection purpose.

Consider avoiding installation unless you can verify the service destinations, what identity and video data are sent, and how credentials and user records are stored. Scanning was partial: one local path-like reference was missing, so the evidence does not cover that referenced item.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
Findings (25)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The supplied chunk contains no surveillance input handling, package detection, inventory processing, or alerting. Database persistence could support a detection skill, but this implementation specifically manages user accounts, personal information, and tokens rather than detection data. Those capabilities are not represented in the declared purpose. The database operations occur when the DAO is instantiated or its methods are called; the chunk does not demonstrate network transmission or automatic execution of those operations on import.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

Generic formatting, polling, and filesystem helpers could reasonably support package detection and are not themselves mismatches. However, the identity provisioning, remote account registration, and credential-bearing network behavior introduce material capabilities not represented by the detection-only description or its empty declared permissions. The request helper also accepts absolute URLs without restricting credential attachment to the configured service. These actions occur when the relevant methods are called, not necessarily upon import. This chunk alone does not establish whether package detection exists elsewhere or whether credentials are actually sent to an untrusted destination.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
88% confidence
Finding

The supplied chunk does not substantiate the declared package-detection purpose and instead implements generic skill scaffolding and an incomplete agent-chat wrapper. It does not actually execute the commented-out subprocess command. Imported base classes and services are not provided, so their behavior cannot be assessed; the mismatch finding is limited to the visible code.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
84% confidence
Finding

The code exposes an analysis-history retrieval operation beyond the declared package-detection purpose. Its primary analysis operation is a generic delegation to skill.get_output_analysis; no package-specific detection logic or configuration is visible in this chunk. The delegated implementation might perform package detection, so that absence alone does not establish a mismatch. Video input handling and optional output-file writing are supporting implementation details, not independent mismatches.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
82% confidence
Finding

Generic AI-analysis endpoints could support the declared detection workflow and are not inherently mismatched. However, DETAIL_EXPORT_URL explicitly targets a health-service order report export resource, which is inconsistent with the declared package-detection purpose. This finding concerns the configured resource; the supplied chunk does not demonstrate that reports are actually accessed or exported.

Content

No source excerpt is available for this finding.

The request helper attaches service credentials to unrestricted absolute destination URLs.

High
Category
Excessive Agency
Confidence
99% confidence
Finding

The helper accepts arbitrary absolute HTTP or HTTPS URLs and automatically adds service tokens and the API key without validating the destination origin. If an attacker can influence a caller's URL, credentials can be sent directly to an attacker-controlled server; accepting HTTP also permits plaintext credential transmission.

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.

The analysis workflow lacks a privacy warning about sending surveillance inputs to a remote API.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

Although the documentation mentions APIs and cloud reports, it does not clearly warn that local surveillance media is uploaded to a remote service or explain its handling. Automatic processing of entrance and lobby footage can disclose faces, delivery labels, addresses, and activity patterns without an informed authorization decision.

Content

No source excerpt is available for this finding.

The --export-env-only option promises not to run analysis, but detection still executes.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The export-only flag is parsed but never checked, so supplying it with an input still invokes detection, and supplying it with --list still retrieves history. This violates the advertised no-analysis boundary and can cause unintended remote processing of sensitive surveillance media.

Content

No source excerpt is available for this finding.

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · skills/smyx_analysis/scripts/skill.py (reported line 28)May include surrounding context.

python
result_json = JsonUtil.parse(result_json_pure_text, result_json_pure_text)

        result_json_common_ai_response = result_json.get("commonAiResponse") if isinstance(result_json,
                                                                                           dict) else result_json
        if result_json_common_ai_response:
            result_json = result_json_common_ai_response

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · skills/smyx_analysis/scripts/skill.py (reported line 33)May include surrounding context.

python
result_json = result_json_common_ai_response

        result_json_health_ai_response = result_json.get("healthAiResponse") if isinstance(result_json,
                                                                                           dict) else result_json
        if result_json_health_ai_response:
            result_json = result_json_health_ai_response

Local video contents are submitted to an analysis API without disclosing the upload.

Medium
Category
Output Handling
Confidence
91% confidence
Finding

The implementation reads the complete local media file and passes it as an upload to the analysis API without a visible disclosure or authorization check. Surveillance footage can include identifiable people and delivery information, making an unexpected remote transfer materially privacy-sensitive.

Content

No source excerpt is available for this finding.

The documented export-only mode still executes video analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The export-only option promises not to analyze media, but no branch implements it before identity resolution, history retrieval, or video analysis. A caller relying on that promise can unintentionally trigger remote processing of private footage.

Content

No source excerpt is available for this finding.

The skill includes user-profile and credential management unrelated to its declared package-detection purpose.

Medium
Category
Excessive Agency
Confidence
88% confidence
Finding

The User model stores real names, email addresses, birthdays, demographic information, and token/open_token credentials, while UserDao exposes inherited read, write, update, and delete operations for these records (L221-L427). DAO initialization also creates tables and migrates sys_user (L180-L215), extending the capability beyond package inventory and unattended alerts into persistent identity and credential management.

Content

No source excerpt is available for this finding.

FileUtil.open silently truncates existing files without disclosing its destructive write mode.

Medium
Category
Output Handling
Confidence
97% confidence
Finding

The generically named open helper opens the supplied path in 'w' mode, immediately erasing existing contents even before the caller writes anything. No docstring, comment, confirmation, or operation-specific log explains this behavior; exception tracing only addresses failures.

Content

No source excerpt is available for this finding.

Package-detection requests can silently register a remote account and persist its credentials.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

An ordinary request can initiate silent registration using the current identity and then persist returned credentials through UserDao, without a consent gate in this flow. This creates remote account state and local credential retention beyond the obvious package-detection operation; the shared-library context explains the functionality but does not remove the consent risk.

Content

No source excerpt is available for this finding.

HTTP requests can silently trigger remote account registration using the current identity.

Medium
Category
Output Handling
Confidence
95% confidence
Finding

The request helper can post a payload containing silent=1 and register=1 to the phoneLogin endpoint as an automatic authentication fallback. Thus an apparently routine package-detection request can create an account using the current identity without an explicit registration decision in the shown flow.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

This POST transmits the current identity as both openId and mobile, together with registration flags and source metadata, during an implicit authentication fallback. External communication can be legitimate for an API-backed detection skill, but the demonstrated issue is undisclosed identity transmission and registration, not evidence that the configured service is malicious.

Content

Scanner excerpt · skills/smyx_common/scripts/util.py (reported line 561)May include surrounding context.

python
"source": ConstantEnum.DEFAULT__SKILL_HUB_NAME
            }
            try:
                _response = requests.post(_url, json=_data)
                if _response.status_code == 200:
                    _response_json = _response.json()
                    if _response_json and _response_json.get("success"):

Loading configuration silently creates directories and writes a file when the configuration is missing.

Low
Category
Output Handling
Confidence
93% confidence
Finding

YamlUtil.load performs filesystem writes without a prompt, log, docstring, or explanatory comment disclosing this side effect. Because BaseEnum invokes this loader during subclass creation (L053 and L058), importing the module can create configuration files without an explicit save request.

Content

No source excerpt is available for this finding.

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

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

python
username = getattr(model, "username", None)
            if username:
                column_names = self.__model__.__table__.columns.keys()
                update_data = {key: getattr(model, key) for key in column_names if key != "username"}
                return self.update_by_username(username, **update_data)

            return None

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/dao.py (reported line 353)May include surrounding context.

python
column_names = self.__model__.__table__.columns.keys()

            for key in column_names:
                value = getattr(model, key)
                setattr(instance, key, value)

            session.commit()

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/dao.py (reported line 326)May include surrounding context.

python
if filters:
                for key, value in filters.items():
                    query = query.filter(getattr(self.__model__, key) == value)

            if offset:
                query = query.offset(offset)

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/dao.py (reported line 441)May include surrounding context.

python
if filters:
                for key, value in filters.items():
                    query = query.filter(getattr(self.__model__, key) == value)

            return query.scalar()
        finally:

ai_chat documents agent execution and analysis results, but its implementation never invokes the agent and returns no result.

Low
Category
Excessive Agency
Confidence
99% confidence
Finding

The docstring at L25–L35 claims the method calls the openclaw agent through subprocess and returns an analysis result or session ID. However, subprocess execution is commented out at L61–L66, and accessing result.stderr on the empty dictionary created at L68 always raises an exception that is swallowed at L81–L82, causing the method to return None without performing analysis.

Content

No source excerpt is available for this finding.

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/util.py (reported line 30)May include surrounding context.

python
elif isinstance(cur, dict):
                        cur = cur.get(part, default)
                    else:
                        cur = getattr(cur, part)
                return cur
            except Exception:
                return default