Back to skill

Security audit

Anxiety-Related Behavior Recognition (Hand-rubbing / Nail-biting / Pacing) | 焦虑症相关行为(搓手、咬指甲、来回踱步)识别

Security checks for vulnerabilities and agentic risk

Overview

The skill is rated suspicious because scanners report sensitive video analysis and data handling alongside account, identity, and credential features outside its stated purpose; the scan is partial because one referenced path could not be resolved.

Consider avoiding installation unless you specifically need this skill and have reviewed its remote processing and account setup. The scanners report potential transfer of sensitive video to an API, identity transmission during automatic registration, and local storage of credentials; the target and retention details are not provided. Coverage is incomplete because a referenced path could not be resolved, so the scan cannot settle behavior associated with that reference.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
Findings (23)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied code does not implement the declared primary purpose. Its only substantive method concerns agent-chat invocation, but it never executes a command because that code is commented out. No undeclared command execution or data transmission is demonstrated. Inherited implementations are not supplied, so their behavior cannot be assessed; this 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
96% confidence
Finding

Generic formatting, polling, and filesystem helpers could reasonably support the declared system. However, this chunk also implements a substantive external account, authentication, and billing integration absent from the description. No camera ingestion, behavior recognition, anxiety scoring, or reminder logic appears in the supplied code. The absence of those features in a utility module alone would not establish a mismatch; the affirmative evidence is the undeclared identity access, remote registration, and account-management behavior. The configured endpoint values and calling code are unavailable, so this assessment does not imply that video is uploaded or that these functions execute merely upon import.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding

The supplied chunk is a generic user-management persistence module rather than a video-analysis or anxiety-monitoring implementation. Database persistence could support the declared system, but this code specifically manages identity and token records without any visible connection to camera input, detected behaviors, anxiety scores, or reminders. That substantive account-management capability is not represented in the description. No network transmission or token exfiltration is shown, and this assessment is limited to the supplied chunk.

Content

No source excerpt is available for this finding.

YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]

High
Category
YARA Match
Confidence
80% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · SKILL.md (reported line 3)May include surrounding context.

md
---
name: "smyx-anxiety-behavior-recognition-analysis"
description: "Using a fixed camera at home or in the office, the system analyzes daily videos of an individual (e.g., adult, adolescent) and detects anxiety-related behaviors: hand rubbing (repeated rubbing of both hands), nail biting (hand approaching mouth with biting motion), and pacing (repeated back-and-forth walking in a small area). | 通过家庭或办公室的固定摄像头,分析个体(如成人、青少年)的日常行为视频,检测手部搓揉(双手反复摩擦)、指甲啃咬(手部靠近嘴部并有啃咬动作)、来回踱步(在狭小区域内反复折返�

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.

Default activation treats any home/office fixed-camera video analysis request as a request for anxiety-behavior assessment.

Medium
Category
Trigger Abuse
Confidence
96% confidence
Finding

The default trigger expands a generic home or office video-analysis request into mental-health-related behavioral profiling without requiring that specific intent. This can produce sensitive inferences about adults or adolescents outside the user's requested scope; the later consent note does not resolve the contradictory automatic activation rule.

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 parser promises that --export-env-only will not perform analysis, but the parsed flag is never checked. With an input or URL present, execution still initializes identity and invokes the normal analysis path, potentially submitting private home or office footage when the caller explicitly selected a non-analysis mode.

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 passed to an API service without disclosure of the upload.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The function reads local video bytes and passes them to the API analysis method without an upload disclosure or consent gate in this path. Home, workplace, or counseling footage can contain highly sensitive personal information, making an unexpected transfer materially more dangerous than ordinary file processing.

Content

No source excerpt is available for this finding.

The --export-env-only option promises not to analyze video, but does not prevent analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The --export-env-only flag is accepted but never used to stop execution. When media input is supplied, the program initializes identity and invokes API analysis despite the explicit promise not to analyze, risking unintended processing or transfer of sensitive footage.

Content

No source excerpt is available for this finding.

The save() method is documented as creating a new record, but insertion failures trigger updates to existing records, including a username-based overwrite fallback (L221–L244).

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

Any insertion exception changes this operation into an update by supplied ID, followed by a username-or-realname overwrite fallback that copies all remaining columns. If an untrusted or misassociated model reaches save(), an existing user's identifiers, tokens, or personal information can be overwritten; sensitive behavior-monitoring use increases the consequences of account misassociation.

Content

No source excerpt is available for this finding.

FileUtil.open silently truncates existing files without disclosing its write-only behavior.

Medium
Category
Output Handling
Confidence
97% confidence
Finding

At L101, opening a path with mode 'w' immediately truncates any existing file. The generically named helper has no docstring, comment, confirmation, or operation log explaining this destructive behavior; exception reporting does not disclose successful truncation.

Content

No source excerpt is available for this finding.

The request helper silently registers an external account and persists its credentials, beyond the declared behavior-analysis and alerting scope.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

An ordinary request can silently send the current identity to an external login endpoint with registration enabled, then persist the returned credentials. The provided code contains no explicit enrollment approval, making this an undisclosed external side effect in a privacy-sensitive home or office monitoring skill; these lines do not establish transmission of video.

Content

No source excerpt is available for this finding.

HTTP requests can silently trigger remote registration and persist returned credentials.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The request helper enables silent registration as a hidden prerequisite to an otherwise ordinary HTTP operation, with returned tokens persisted later in the same flow. This bypasses an explicit account-enrollment decision and creates unexpected identity linkage, particularly concerning for anxiety-related monitoring.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
92% confidence
Finding

The POST sends the selected identity as both openId and mobile, plus the skill-hub source, to a configured external service without an approval check in this flow. The identity may be generated or user-derived, so personal-information exposure depends on configuration, but the undisclosed external registration is concrete.

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"):

Configuration loading silently creates directories and writes missing configuration files.

Low
Category
Output Handling
Confidence
94% confidence
Finding

At L017–L020, YamlUtil.load creates parent directories and writes a YAML file when the requested path is missing, without a warning, log, comment, or docstring disclosing this side effect. Because BaseEnum.init_subclass invokes this loader at L053 and L058, merely importing the module can modify the filesystem rather than only read configuration.

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 claims to invoke an analysis agent, but its invocation is disabled and it returns no analysis.

Low
Category
Excessive Agency
Confidence
99% confidence
Finding

The docstring at L25–L34 explicitly promises to invoke openclaw agent through subprocess and return an analysis result or session ID. However, the subprocess call is commented out, and L68–L70 instead create an empty dictionary and access its nonexistent stderr attribute; the exception is caught, causing the method to return None without invoking an agent.

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