Back to skill

Security audit

Workplace Group Stress Heatmap | 职场员工压力群体热力图

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious with high confidence: scanners report that video handling and account-related behavior raise privacy and scope concerns for the described anonymous workplace stress analysis.

Treat this as a review item and weigh the reported workplace-video upload, trigger behavior, destination handling, and account registration before installing. Scanner coverage is partial because a referenced local path was missing, so the available evidence does not confirm the artifact's full behavior.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/smyx_workplace_stress_heatmap_analysis.py
Finding

Ignored API destination override sends sensitive media to the default provider

Content
View full analysis

Impact

Identifiable workplace footage or private media URLs can be disclosed to a recipient the user expressly attempted to replace. File-size and extension validation do not constrain the recipient.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Findings (25)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The visible implementation does not substantiate the declared purpose and has a materially different focus. There is no active subprocess execution or evidence of camera access in this chunk. Imported base classes and services are not supplied, so their behavior cannot be assessed; the conclusion is limited to the visible code.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The supplied chunk contains no camera processing, expression or posture analysis, group stress model, heatmap generation, or periodic reporting. Database persistence could be a supporting implementation detail, but this code specifically creates and manages a sys_user table containing identity and token fields rather than anonymous zone-level stress data. That is a material capability not represented by the declared purpose. This assessment is limited to the supplied chunk; it does not establish that employee identities are linked to camera observations or that data is transmitted externally.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

Ordinary formatting, filesystem, and polling helpers could support the declared system and are not themselves evidence of a mismatch. However, this chunk additionally implements an undeclared account-identity, remote-registration, authentication, and billing workflow. It contains no camera ingestion, expression or posture analysis, stress-assessment model, anonymization, or heatmap generation. Persistent account identity does not by itself prove that employees are identified, and this shared utility file cannot establish the behavior of unseen code, but its substantial account-service capabilities are not represented in the supplied description.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

API submission and result polling could support the declared purpose, but this chunk exposes only a generic video-analysis/reporting client. It neither implements nor explicitly requests the office-specific stress assessment described. The imported API implementation is not supplied, so those capabilities might exist remotely; however, the visible code does not substantiate the declared specialized behavior. No malicious destination or unrelated data theft is established.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
86% confidence
Finding

The supplied code handles video inputs and calls skill.get_output_analysis(), but contains no visible facial-expression or posture processing, group stress scoring, workstation-zone mapping, heatmap generation, or periodic execution. It also resolves an Open ID and offers a separate history-listing operation through skill.get_output_analysis_list(), an undeclared capability. The imported skill implementation is unavailable, so specialized stress analysis could occur there; its presence and anonymization guarantees cannot be established from this chunk. Open ID usage alone does not establish that employees are identified.

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 default trigger selects stress analysis for any office-camera video analysis request, even without a request to assess stress.

Medium
Category
Trigger Abuse
Confidence
98% confidence
Finding

The default trigger expands any office-camera analysis request into employee stress assessment without requiring that purpose to be requested. In a workplace, this can initiate sensitive behavioral inference and cloud processing outside the user's intended task; general deployment-consent advice does not constrain the trigger.

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 advertises an export-only option that performs no analysis, but the parsed value is never checked. When supplied with valid input, the command still initializes identity and submits video for analysis, potentially exposing sensitive office footage despite an explicit no-analysis choice.

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
95% confidence
Finding

The implementation reads the original local video and passes its bytes to the analysis service without local anonymization or an upload-confirmation gate. Identifiable employee footage therefore crosses a trust boundary even though users may interpret the anonymity claims as protecting the input, rather than only the resulting statistics.

Content

No source excerpt is available for this finding.

The documented export-only mode does not prevent video analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The export-only argument promises not to analyze video, but execution never branches on its value. With a valid input, analysis proceeds normally, defeating the caller's explicit restriction and potentially transmitting confidential workplace video.

Content

No source excerpt is available for this finding.

Username-based user lookup is outside the scope of anonymous, group-level stress analysis.

Medium
Category
Excessive Agency
Confidence
91% confidence
Finding

The manifest describes anonymous analysis of employees to produce zone-level stress heatmaps. This method instead exposes user lookup by username, an identity-oriented capability with no stated role in that workflow; the code does not establish that it is used to identify people in camera footage.

Content

No source excerpt is available for this finding.

Identifiable-user and credential management is unrelated to anonymous, zone-level stress analysis.

Medium
Category
Excessive Agency
Confidence
95% confidence
Finding

The User model stores usernames, real names, email addresses, demographic details, and token/open_token values, and UserDao exposes the inherited database CRUD operations for these records. DAO initialization also creates and migrates the user table (L184–L215). This implements persistent identity and credential management beyond the stated anonymous group-level analytics purpose; the file does not establish that employee identities are actually collected or linked to stress results.

Content

No source excerpt is available for this finding.

FileUtil.open silently truncates existing files.

Medium
Category
Output Handling
Confidence
97% confidence
Finding

The generic open helper always uses mode 'w', which immediately erases existing file contents. No docstring, comment, confirmation, or action log discloses this destructive behavior; the exception handler only runs after a failure.

Content

No source excerpt is available for this finding.

Analytics requests can silently register a remote user account and persist its authentication credentials.

Medium
Category
Excessive Agency
Confidence
97% confidence
Finding

An ordinary HTTP request can initiate silent remote registration and persist the returned authentication tokens through UserDao without a separate enrollment decision. For an anonymously described workplace analytics skill, this introduces undisclosed identity transmission and credential retention, although employee-level identification is not demonstrated.

Content

No source excerpt is available for this finding.

HTTP requests can silently trigger remote account registration without disclosure.

Medium
Category
Output Handling
Confidence
98% confidence
Finding

The request helper submits silent=1 and register=1 to a remote login endpoint when credentials are missing, making account creation an implicit side effect of a normal request. This bypasses an explicit enrollment decision and can disclose the selected identity to an external service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
96% confidence
Finding

This POST sends the selected identity as both openId and mobile without a separate disclosure or consent step. The caller prioritizes API_SECRET_KEY as that identity, so a configured API secret can also be copied into registration fields, potentially exposing it to identity records and logs beyond its intended authentication use.

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 missing configuration files, including during module import.

Low
Category
Output Handling
Confidence
90% confidence
Finding

YamlUtil.load creates parent directories and writes a YAML file when the requested path is absent, without a comment, docstring, prompt, or log disclosing this side effect. BaseEnum.init_subclass invokes this loader at L053 and L058, so 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 documents an OpenClaw agent invocation but never executes it.

Low
Category
Excessive Agency
Confidence
99% confidence
Finding

The docstring at L26 explicitly states that this method invokes openclaw agent through subprocess. However, command construction and execution are commented out, and the empty dictionary assigned at L68 causes the attribute access at L70 to fail; the exception is caught at L81–L82, so the method returns no analysis result without ever invoking the 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