Back to skill

Security audit

Pregnancy Emotion Soothing | 孕妇情绪波动舒缓

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious because scanners report a mismatch between its advertised pregnancy-emotion monitoring and its reported on-demand video analysis, along with privacy-sensitive uploads and account setup that may happen without clear disclosure.

Consider avoiding installation unless you are comfortable with the reported external video processing and automatic account setup. Before using it with sensitive recordings, verify who receives the video and identity details, what is retained, and whether upload and enrollment can be explicitly authorized or avoided. The scan was partial because a referenced local path was missing, so the evidence does not establish what that missing reference would contain.

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
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
Findings (24)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The visible implementation materially differs from the declared purpose: it provides generic skill scaffolding and a nonfunctional chat wrapper, not the described emotional monitoring and automated interventions. No agent subprocess actually executes, and no camera, microphone, speaker, or notification access is visible. Imported base classes could contain additional behavior, but that cannot be established from this chunk.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied chunk implements account and API infrastructure rather than the described emotion-monitoring workflow. Generic formatting, polling, and HTTP helpers could support that workflow and are not independently mismatches. However, automatic identity provisioning, remote registration, persistent authentication, and payment-related behavior introduce substantial undeclared functionality. No camera or microphone processing, inactivity detection, soothing-audio control, or husband notification is implemented in this chunk. This assessment is limited to the supplied code; it does not establish what other skill files implement.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied chunk's primary function is general user-database management rather than the declared emotional-monitoring workflow. Database storage could support a larger application, but this code contains no emotion-related records or integration linking its user and token management to the stated purpose. The mismatch is limited to the supplied chunk; it does not establish malicious behavior or show what other files implement.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The skill promises continuous monitoring and automatic interventions, but the supplied workflow demonstrates on-demand API analysis and report retrieval without verified intervention delivery. In this pregnancy mental-health context, unsupported assurances can cause users to rely on monitoring or alerts that are not actually operating; historical report retrieval itself is documented and is not suspicious.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding

The visible implementation provides an on-demand analysis workflow, whereas the description presents an active monitoring and notification service. This mismatch creates potentially dangerous reliance in a health-related setting, although the separate claim that history retrieval is undisclosed is contradicted by the skill's detailed documentation.

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-pregnancy-emotion-soothing-analysis"
description: "Through fixed cameras (and optional microphones) at the pregnant woman's home or prenatal exam waiting room, the system analyzes facial expressions (sudden crying, frowning, anxiety), prolonged silent sitting (≥ 30 consecutive minutes without social interaction or activity), and tone of conversations with family members (rapid, impatient). | 通过孕妇家中或产检候诊室的固定摄像头(及可选麦克风),分析孕妇的面部表情(突然哭泣、皱眉、焦虑)、长时间静坐不语(连续超过30分钟无社交互动或活动)、以及与家人对�

The local-media analysis workflow lacks a clear warning about external processing of sensitive recordings.

High
Category
Output Handling
Confidence
97% confidence
Finding

The local-file workflow does not clearly disclose that private recordings will be uploaded for external analysis or require informed consent before transfer. Home and prenatal waiting-room recordings can expose health information, intimate conversations, and bystanders; general API references and retention promises do not provide adequate upload disclosure.

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 --export-env-only option promises not to analyze media, but execution still invokes analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The parser advertises --export-env-only as not performing analysis, but the flag is never checked and a supplied input still reaches analyze_video. Someone relying on this mode can unintentionally submit sensitive pregnancy-related recordings for remote processing, and identity initialization also occurs before any export-only handling.

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 the analysis API without disclosure that the file will be uploaded.

Medium
Category
Output Handling
Confidence
98% confidence
Finding

The function reads the complete local recording into a multipart file payload and passes it to the analysis service without a disclosure or consent gate. For home and prenatal recordings, this can expose highly sensitive audiovisual information beyond the user's expectation of local-file analysis.

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-env-only argument is accepted but never used to alter control flow, so a command with an input still performs video analysis. This violates the explicit no-analysis assurance and can unintentionally initiate external processing of private recordings; identity resolution also runs regardless of the flag.

Content

No source excerpt is available for this finding.

ai_chat documents an agent subprocess invocation, but the implementation never invokes the agent or produces its analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The docstring at L25–L34 states that this method calls openclaw agent through subprocess and returns an analysis result or session ID. However, the subprocess invocation is commented out, and L68 substitutes an empty dictionary whose stderr attribute access raises an exception; the handler then lets the method return None. Callers therefore receive no documented analysis, potentially leaving a dependent emotion-analysis workflow without results.

Content

No source excerpt is available for this finding.

FileUtil.open silently truncates existing files without disclosing its destructive behavior.

Medium
Category
Output Handling
Confidence
96% confidence
Finding

The generically named open helper always uses mode 'w', which immediately erases an existing file's contents. There is no confirmation, operation log, docstring, or comment explaining this destructive side effect; exception tracing only occurs after a failure.

Content

No source excerpt is available for this finding.

Ordinary API requests can silently register a remote account and persist its credentials, beyond the declared emotion-monitoring and soothing functions.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

When credentials are missing, an ordinary request can invoke phoneLogin with silent=1 and register=1, transmit the current identity, and save returned tokens through UserDao without an explicit enrollment step. This creates an undisclosed account and credential lifecycle; the pregnancy-monitoring context makes unexpected identity linkage particularly sensitive, although this snippet does not demonstrate transmission of recordings or emotion data.

Content

No source excerpt is available for this finding.

Ordinary HTTP requests can silently trigger remote account registration using the current user identifier.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The request helper can silently register an account using the current user identifier when tokens are unavailable, rather than limiting itself to the requested HTTP operation. This is the same implicit-enrollment issue identified by SDI-1, not a separate exploit, and can unexpectedly associate a user with a sensitive pregnancy-related service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

This POST sends the current identity as both openId and mobile, together with a source identifier, to the configured health-service endpoint during automatic login or registration. Network access itself is expected for app notifications, but this undisclosed identity transmission is an additional privacy-sensitive side effect; the destination's ownership and receipt of any monitoring data are not established here.

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 files, including during module import (L017–L020).

Low
Category
Output Handling
Confidence
92% confidence
Finding

YamlUtil.load creates a parent directory and writes a YAML file when the requested configuration is absent, without a warning, log message, or explanatory comment/docstring. BaseEnum.init_subclass calls this loader as classes are defined (L044–L059), so merely importing the module can modify the filesystem without disclosure.

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 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:

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/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